Indian address classification.ipynb

{
  "cells": [
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "nQum2GeWiamD"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "fpxiIuaYiRFb"
      },
      "outputs": [],
      "source": [
        "import pandas as pd\n",
        "import numpy as np\n",
        "import torch\n",
        "import re\n",
        "\n",
        "#visualization\n",
        "import seaborn as sns\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "from collections import defaultdict\n",
        "\n",
        "import string, re, nltk\n",
        "from string import punctuation\n",
        "from nltk.tokenize import word_tokenize, RegexpTokenizer\n",
        "from nltk.corpus import stopwords\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 455
        },
        "id": "phpRzy2ziRFd",
        "outputId": "73cf94c6-ebc0-4876-808b-d610919814f7"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                                          masked_address       label\n",
              "id                                                                  \n",
              "2      JAI BALAJI DISTRIBUTORS, NEW MARKET SHOP NO.**...  commercial\n",
              "423    SUNDRAM FASTENERS LIMITED B ** SIPCOT INDUSTRI...  commercial\n",
              "481    Bakul-*, Shardashish Township, Torrent Power L...  commercial\n",
              "641    Sonkar Bhawan, Near Ashok Kirana Store, Infron...  commercial\n",
              "4395   jambli \\n maida factory near jambli  , jambli,...  commercial\n",
              "...                                                  ...         ...\n",
              "10596  A. Rajeshwari, No.***, Pampatti street, East g...         NaN\n",
              "10597  Flat no ***,S.V's srinivasam, nehrunagar  , HY...         NaN\n",
              "10598  Madhav nagar katni, Near post office road  , M...         NaN\n",
              "10599  Mig-**, sri sai -* men s pg, Road no-*, kphb c...         NaN\n",
              "10600  *th Mile Tadong, Near Sikkim Milk Union, New H...         NaN\n",
              "\n",
              "[10601 rows x 2 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-5871da14-5d2b-49dc-b6fc-ee3348156cdd\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>masked_address</th>\n",
              "      <th>label</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>id</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>JAI BALAJI DISTRIBUTORS, NEW MARKET SHOP NO.**...</td>\n",
              "      <td>commercial</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>423</th>\n",
              "      <td>SUNDRAM FASTENERS LIMITED B ** SIPCOT INDUSTRI...</td>\n",
              "      <td>commercial</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>481</th>\n",
              "      <td>Bakul-*, Shardashish Township, Torrent Power L...</td>\n",
              "      <td>commercial</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>641</th>\n",
              "      <td>Sonkar Bhawan, Near Ashok Kirana Store, Infron...</td>\n",
              "      <td>commercial</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4395</th>\n",
              "      <td>jambli \\n maida factory near jambli  , jambli,...</td>\n",
              "      <td>commercial</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10596</th>\n",
              "      <td>A. Rajeshwari, No.***, Pampatti street, East g...</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10597</th>\n",
              "      <td>Flat no ***,S.V's srinivasam, nehrunagar  , HY...</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10598</th>\n",
              "      <td>Madhav nagar katni, Near post office road  , M...</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10599</th>\n",
              "      <td>Mig-**, sri sai -* men s pg, Road no-*, kphb c...</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10600</th>\n",
              "      <td>*th Mile Tadong, Near Sikkim Milk Union, New H...</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>10601 rows × 2 columns</p>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-5871da14-5d2b-49dc-b6fc-ee3348156cdd')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-5871da14-5d2b-49dc-b6fc-ee3348156cdd button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-5871da14-5d2b-49dc-b6fc-ee3348156cdd');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-45303d37-964d-4f85-9073-aab9b7fe7859\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-45303d37-964d-4f85-9073-aab9b7fe7859')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-45303d37-964d-4f85-9073-aab9b7fe7859 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "\n",
              "  <div id=\"id_a45001e6-e182-4e3e-95a6-799260cdc506\">\n",
              "    <style>\n",
              "      .colab-df-generate {\n",
              "        background-color: #E8F0FE;\n",
              "        border: none;\n",
              "        border-radius: 50%;\n",
              "        cursor: pointer;\n",
              "        display: none;\n",
              "        fill: #1967D2;\n",
              "        height: 32px;\n",
              "        padding: 0 0 0 0;\n",
              "        width: 32px;\n",
              "      }\n",
              "\n",
              "      .colab-df-generate:hover {\n",
              "        background-color: #E2EBFA;\n",
              "        box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "        fill: #174EA6;\n",
              "      }\n",
              "\n",
              "      [theme=dark] .colab-df-generate {\n",
              "        background-color: #3B4455;\n",
              "        fill: #D2E3FC;\n",
              "      }\n",
              "\n",
              "      [theme=dark] .colab-df-generate:hover {\n",
              "        background-color: #434B5C;\n",
              "        box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "        filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "        fill: #FFFFFF;\n",
              "      }\n",
              "    </style>\n",
              "    <button class=\"colab-df-generate\" onclick=\"generateWithVariable('df')\"\n",
              "            title=\"Generate code using this dataframe.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M7,19H8.4L18.45,9,17,7.55,7,17.6ZM5,21V16.75L18.45,3.32a2,2,0,0,1,2.83,0l1.4,1.43a1.91,1.91,0,0,1,.58,1.4,1.91,1.91,0,0,1-.58,1.4L9.25,21ZM18.45,9,17,7.55Zm-12,3A5.31,5.31,0,0,0,4.9,8.1,5.31,5.31,0,0,0,1,6.5,5.31,5.31,0,0,0,4.9,4.9,5.31,5.31,0,0,0,6.5,1,5.31,5.31,0,0,0,8.1,4.9,5.31,5.31,0,0,0,12,6.5,5.46,5.46,0,0,0,6.5,12Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "    <script>\n",
              "      (() => {\n",
              "      const buttonEl =\n",
              "        document.querySelector('#id_a45001e6-e182-4e3e-95a6-799260cdc506 button.colab-df-generate');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      buttonEl.onclick = () => {\n",
              "        google.colab.notebook.generateWithVariable('df');\n",
              "      }\n",
              "      })();\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "    </div>\n",
              "  </div>\n"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "dataframe",
              "variable_name": "df",
              "summary": "{\n  \"name\": \"df\",\n  \"rows\": 10601,\n  \"fields\": [\n    {\n      \"column\": \"id\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 3060,\n        \"min\": 0,\n        \"max\": 10600,\n        \"num_unique_values\": 10601,\n        \"samples\": [\n          2735,\n          2067,\n          1451\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"masked_address\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 10601,\n        \"samples\": [\n          \"MORE RETAIL LTD   Survey No **/Part, NH-* Baypass, Nr Sadumalam,   PENDURTHI, GVMC, Vishakapatnam,   Vishakapatnam, AP-  , Pendurthi, Andhra Pradesh, 531173\",\n          \"Ashok NagarFATEHPUR  , FATEHPUR, Uttar Pradesh, 212601\",\n          \"Nainar kovil \\n Thiri.Maruthur  , Nainar kovil, Tamil Nadu, 623705\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"label\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 2,\n        \"samples\": [\n          \"residential\",\n          \"commercial\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}"
            }
          },
          "metadata": {},
          "execution_count": 7
        }
      ],
      "source": [
        "## load the data\n",
        "df=pd.read_csv('/content/assignment_data_res_com.csv',index_col=0)\n",
        "df"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "4ngJPDo1iRFd"
      },
      "outputs": [],
      "source": [
        "## EDA for labelled data\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "woF1vBmyiRFe",
        "outputId": "ad3e783b-5151-4bc4-abdb-cd8e01d4128d"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Number of observations with missing values    0\n",
            "Number of duplicate observations              0\n"
          ]
        }
      ],
      "source": [
        "# Missing values and duplicate address?\n",
        "print(pd.Series({\"Number of observations with missing values\": len(df['masked_address']) - len(df['masked_address'].dropna()),\n",
        "                 \"Number of duplicate observations\": df['masked_address'].duplicated().sum()}).to_string())"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "4xmtGmgQiRFe"
      },
      "outputs": [],
      "source": [
        "# Manual encoding of labels\n",
        "label_dict = {'commercial': 0, 'residential': 1}\n",
        "df.replace({'label': label_dict}, inplace = True)\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "MG7Ezk7wiRFe"
      },
      "outputs": [],
      "source": [
        "# Splitting the dataset by label\n",
        "data_c = df[df['label'] == 0] # commerical\n",
        "data_r = df[df['label'] == 1] # residential"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 433
        },
        "id": "42K8tIT_iRFe",
        "outputId": "a721a953-d541-4a15-ce77-abe015b372e1"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x420 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "# Distribution of number of words in description\n",
        "data_c_word = data_c['masked_address'].str.split().map(lambda x: len(x))\n",
        "data_r_word = data_r['masked_address'].str.split().map(lambda x: len(x))\n",
        "\n",
        "fig, ax = plt.subplots(1, 2, figsize=(10, 4.2), sharey=False)\n",
        "sns.histplot(x=data_c_word, bins=20, ax=ax[0]).set_title('Class: commercial')\n",
        "sns.histplot(x=data_r_word, bins=20, ax=ax[1]).set_title('Class: residential')\n",
        "\n",
        "fig.suptitle(\"Distribution of number of words in description\")\n",
        "\n",
        "for i in range(2):\n",
        "    ax[i].set_xlabel(\"Number of words\")\n",
        "    if i != 0:\n",
        "        ax[i].set_ylabel(\" \")\n",
        "\n",
        "plt.tight_layout()\n",
        "plt.show()\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "IRZVMNAciRFf"
      },
      "outputs": [],
      "source": [
        "## Text Normalisation/Pre-processing\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "TQ0ALLMaiRFf"
      },
      "outputs": [],
      "source": [
        "# Converting to lowercase\n",
        "df['masked_address'] = df['masked_address'].str.lower()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "Q3UrGyr9iRFf"
      },
      "outputs": [],
      "source": [
        "## remove emails\n",
        "def remove_custom_emails(text):\n",
        "    # Pattern to remove emails with various masking\n",
        "    custom_email_pattern = r'[A-Za-z0-9.*_%-]+@(?:gmail\\.com|yahoo\\.com|hotmail\\.com|outlook\\.com)'\n",
        "    return re.sub(custom_email_pattern, '', text)\n",
        "\n",
        "\n",
        "df['masked_address']=df['masked_address'].apply(remove_custom_emails)\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "9rIb7SvSiRFf"
      },
      "outputs": [],
      "source": [
        "\n",
        "# Remove all special characters except * and $ (masks) and /\n",
        "\n",
        "def clean_special_characters(text):\n",
        "    return re.sub(r'[^\\w\\s*$/]', ' ', text)\n",
        "\n",
        "df['masked_address']=df['masked_address'].apply(clean_special_characters)"
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "1Tz6dlrJ1EWl"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "n5ld9-mTiRFf"
      },
      "outputs": [],
      "source": [
        "# Function to normalize spacing\n",
        "def normalize_spacing(text):\n",
        "    # Replace multiple spaces with a single space\n",
        "    return re.sub(r'\\s+', ' ', text).strip()\n",
        "\n",
        "\n",
        "df['masked_address'] = df['masked_address'].apply(normalize_spacing)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "xBws_ZtdiRFg"
      },
      "outputs": [],
      "source": [
        "# Function to remove consecutive duplicate words\n",
        "def remove_consecutive_duplicates(text):\n",
        "    # Regex pattern to match consecutive duplicate words\n",
        "    pattern = r'\\b(\\w+)\\s+\\1\\b'\n",
        "    # Replace consecutive duplicate words with a single instance\n",
        "    return re.sub(pattern, r'\\1', text)\n",
        "\n",
        "df['masked_address'] = df['masked_address'].apply(remove_consecutive_duplicates)\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "Os0DdvBjiRFg"
      },
      "outputs": [],
      "source": [
        "# Function to remove all occurrences of duplicate words irrespective of consecutive or not\n",
        "def remove_all_duplicates(text):\n",
        "    words = text.split()\n",
        "    seen = set()\n",
        "    result = []\n",
        "    for word in words:\n",
        "        # Check if the word is already in the seen set\n",
        "        if word.lower() not in seen:\n",
        "            seen.add(word.lower())\n",
        "            result.append(word)\n",
        "    return ' '.join(result)\n",
        "\n",
        "df['masked_address'] = df['masked_address'].apply(remove_all_duplicates)\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "iRFkLJz9iRFg"
      },
      "outputs": [],
      "source": [
        "#df.to_csv('c:/Users/Anjali/Downloads/Assignment/assignment_cleaned.csv')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "NVwmPogPiRFg"
      },
      "outputs": [],
      "source": [
        "# ## Probabilistic splitting\n",
        "# def probabilistic_split(token, term_freq_dict):\n",
        "#     # Check if the token is masked\n",
        "#     if '$' in token or '*' in token:\n",
        "#         return token, None\n",
        "\n",
        "#     n = len(token)\n",
        "#     for i in range(1, n):\n",
        "#         token1, token2 = token[:i], token[i:]\n",
        "#         # Check if the split tokens have a higher combined frequency\n",
        "#         if term_freq_dict[token1] > 0 and term_freq_dict[token2] > 0:\n",
        "#             if term_freq_dict[token1] + term_freq_dict[token2] > term_freq_dict[token]:\n",
        "#                 return token1, token2\n",
        "#     return token, None\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "6Pb87tlqiRFg"
      },
      "outputs": [],
      "source": [
        "# Create term-frequency dictionary\n",
        "term_freq_dict = defaultdict(int)\n",
        "for address in df['masked_address']:\n",
        "    tokens = address.split()\n",
        "    for token in tokens:\n",
        "        term_freq_dict[token] += 1"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "6tPFWPAOiRFg"
      },
      "outputs": [],
      "source": [
        "\n",
        "# # Split each address based on probabilistic splitting\n",
        "# def split_address(address, term_freq_dict):\n",
        "#     tokens = address.split()\n",
        "#     split_tokens = []\n",
        "#     for token in tokens:\n",
        "#         token1, token2 = probabilistic_split(token, term_freq_dict)\n",
        "#         split_tokens.append(token1)\n",
        "#         if token2:\n",
        "#             split_tokens.append(token2)\n",
        "#     return ' '.join(split_tokens)\n",
        "\n",
        "\n",
        "# df['processed_address'] = df['masked_address'].apply(lambda x: split_address(x, term_freq_dict))\n",
        "\n",
        "# print(df[['masked_address', 'processed_address']])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "xHQC4gb4iRFg"
      },
      "outputs": [],
      "source": [
        "# # pyspellchecker\n",
        "# spell = SpellChecker()\n",
        "\n",
        "# def pyspellchecker(text):\n",
        "#     word_list = regexp.tokenize(text)\n",
        "#     word_list_corrected = []\n",
        "#     for word in word_list:\n",
        "#         if word in spell.unknown(word_list):\n",
        "#             word_corrected = spell.correction(word)\n",
        "#             if word_corrected == None:\n",
        "#                 word_list_corrected.append(word)\n",
        "#             else:\n",
        "#                 word_list_corrected.append(word_corrected)\n",
        "#         else:\n",
        "#             word_list_corrected.append(word)\n",
        "#     text_corrected = \" \".join(word_list_corrected)\n",
        "#     return text_corrected\n",
        "\n",
        "# text = \"Rajpur Road dehradun,uttrakhand\"\n",
        "# print(\"Input: {}\".format(text))\n",
        "# print(\"Output: {}\".format(pyspellchecker(text)))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "ZOmgexPgiRFg"
      },
      "outputs": [],
      "source": [
        "#df.to_csv('c:/Users/Anjali/Downloads/Assignment/df_corrected_spell.csv')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "rRhahPI_iRFh",
        "outputId": "772aee2e-5a73-4993-aedb-6b30b884c477"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/sklearn/feature_extraction/text.py:525: UserWarning: The parameter 'token_pattern' will not be used since 'tokenizer' is not None'\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "                                          masked_address  \\\n",
            "id                                                         \n",
            "2      jai balaji distributors new market shop no ** ...   \n",
            "423    sundram fasteners limited b ** sipcot industri...   \n",
            "481    bakul * shardashish township torrent power lim...   \n",
            "641    sonkar bhawan near ashok kirana store infront ...   \n",
            "4395        jambli maida factory near maharashtra 416121   \n",
            "...                                                  ...   \n",
            "10596  a rajeshwari no *** pampatti street east gate ...   \n",
            "10597  flat no *** s v srinivasam nehrunagar hyderaba...   \n",
            "10598  madhav nagar katni near post office road murwa...   \n",
            "10599  mig ** sri sai * men s pg road no kphb colony ...   \n",
            "10600  *th mile tadong near sikkim milk union new hot...   \n",
            "\n",
            "                                          address_vector  \n",
            "id                                                        \n",
            "2      [-0.06931223549757207, 0.11700991654848282, 0....  \n",
            "423    [-0.0619900122260751, 0.14758910422086188, -0....  \n",
            "481    [-0.07083080284310861, 0.14458166870855566, -0...  \n",
            "641    [-0.07039106557285, 0.14437521573040477, -0.00...  \n",
            "4395   [-0.06386423349870973, 0.14849421964728765, 0....  \n",
            "...                                                  ...  \n",
            "10596  [-0.06445098798100472, 0.14036478529048133, -0...  \n",
            "10597  [-0.07216214826157326, 0.1010708390363707, 0.0...  \n",
            "10598  [-0.07936569826737797, 0.15118979964838652, -0...  \n",
            "10599  [-0.07299363993377773, 0.12335054816609153, 0....  \n",
            "10600  [-0.06328782006464252, 0.12171219619961539, 0....  \n",
            "\n",
            "[10601 rows x 2 columns]\n"
          ]
        }
      ],
      "source": [
        "### tokenisation with WOrd2vec and weighing with TF-IDF\n",
        "\n",
        "## Tokens and vector of the tokenised address\n",
        "#The final vector is normalized to unit length of 100 which help in training the algorithm.\n",
        "import re\n",
        "import pandas as pd\n",
        "from sklearn.feature_extraction.text import TfidfVectorizer\n",
        "from gensim.models import Word2Vec\n",
        "\n",
        "# Step 1: Tokenize addresses\n",
        "def tokenize_address(address):\n",
        "    # Keep PIN codes as single tokens\n",
        "    address = re.sub(r'(\\d{6})', r' \\1 ', address)\n",
        "    # Tokenize the rest while keeping * and $\n",
        "    #tokens = re.findall(r'\\*\\*+|\\$\\$+|\\w+', address)\n",
        "    ## can better it later idea is to preserve * and $ associated text and / represents some number fromat\n",
        "    tokens = re.findall(r'c\\/o|\\*\\*+|\\$\\$+|\\w+\\/\\w+|\\w+\\/\\*+|\\*+\\/\\*+|\\*+\\/\\w+|\\w+', address)\n",
        "    return tokens\n",
        "\n",
        "# def tokenize_address(address):\n",
        "#     # Updated regex pattern to match **, $$, c/o, and sequences with /\n",
        "#     tokens = re.findall(r'\\*\\*\\/\\*|\\$\\$+|c\\/o|\\*+\\/\\*+|\\w+', address)\n",
        "#     return tokens\n",
        "\n",
        "df['tokens'] = df['masked_address'].apply(tokenize_address)\n",
        "\n",
        "# Step 2: Compute TF-IDF weights\n",
        "corpus = df['masked_address']\n",
        "vectorizer = TfidfVectorizer(tokenizer=tokenize_address)\n",
        "X = vectorizer.fit_transform(corpus)\n",
        "tfidf_dict = dict(zip(vectorizer.get_feature_names_out(), vectorizer.idf_))\n",
        "\n",
        "# Step 3: Train Word2Vec model\n",
        "tokens = df['tokens'].tolist()\n",
        "w2v_model = Word2Vec(sentences=tokens, vector_size=100, window=5, min_count=1, workers=4)\n",
        "\n",
        "# Step 4: Generate address vectors using TF-IDF\n",
        "\n",
        "\n",
        "def get_weighted_avg_vector(tokens, tfidf_dict, w2v_model):\n",
        "    vector_sum = np.zeros(w2v_model.vector_size)\n",
        "    total_weight = 0\n",
        "    for token in tokens:\n",
        "        if token in w2v_model.wv:\n",
        "            weight = tfidf_dict.get(token, 1.0)  # Default weight 1.0 for unseen tokens\n",
        "            vector_sum += w2v_model.wv[token] * weight\n",
        "            total_weight += weight\n",
        "    if total_weight == 0:\n",
        "        return np.zeros(w2v_model.vector_size)  # Return zero vector instead of None\n",
        "    weighted_vector = vector_sum / total_weight\n",
        "    return weighted_vector / np.linalg.norm(weighted_vector)  # Normalize to unit length\n",
        "\n",
        "df['address_vector'] = df['tokens'].apply(lambda x: get_weighted_avg_vector(x, tfidf_dict, w2v_model))\n",
        "\n",
        "print(df[['masked_address', 'address_vector']])\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 455
        },
        "id": "OmCobb2riRFh",
        "outputId": "a77bf11d-5114-43ee-e268-dea21938d301"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                                          masked_address  label  \\\n",
              "id                                                                \n",
              "2      jai balaji distributors new market shop no ** ...    0.0   \n",
              "423    sundram fasteners limited b ** sipcot industri...    0.0   \n",
              "481    bakul * shardashish township torrent power lim...    0.0   \n",
              "641    sonkar bhawan near ashok kirana store infront ...    0.0   \n",
              "4395        jambli maida factory near maharashtra 416121    0.0   \n",
              "...                                                  ...    ...   \n",
              "10596  a rajeshwari no *** pampatti street east gate ...    NaN   \n",
              "10597  flat no *** s v srinivasam nehrunagar hyderaba...    NaN   \n",
              "10598  madhav nagar katni near post office road murwa...    NaN   \n",
              "10599  mig ** sri sai * men s pg road no kphb colony ...    NaN   \n",
              "10600  *th mile tadong near sikkim milk union new hot...    NaN   \n",
              "\n",
              "                                                  tokens  \\\n",
              "id                                                         \n",
              "2      [jai, balaji, distributors, new, market, shop,...   \n",
              "423    [sundram, fasteners, limited, b, **, sipcot, i...   \n",
              "481    [bakul, shardashish, township, torrent, power,...   \n",
              "641    [sonkar, bhawan, near, ashok, kirana, store, i...   \n",
              "4395   [jambli, maida, factory, near, maharashtra, 41...   \n",
              "...                                                  ...   \n",
              "10596  [a, rajeshwari, no, ***, pampatti, street, eas...   \n",
              "10597  [flat, no, ***, s, v, srinivasam, nehrunagar, ...   \n",
              "10598  [madhav, nagar, katni, near, post, office, roa...   \n",
              "10599  [mig, **, sri, sai, men, s, pg, road, no, kphb...   \n",
              "10600  [th, mile, tadong, near, sikkim, milk, union, ...   \n",
              "\n",
              "                                          address_vector  \n",
              "id                                                        \n",
              "2      [-0.06931223549757207, 0.11700991654848282, 0....  \n",
              "423    [-0.0619900122260751, 0.14758910422086188, -0....  \n",
              "481    [-0.07083080284310861, 0.14458166870855566, -0...  \n",
              "641    [-0.07039106557285, 0.14437521573040477, -0.00...  \n",
              "4395   [-0.06386423349870973, 0.14849421964728765, 0....  \n",
              "...                                                  ...  \n",
              "10596  [-0.06445098798100472, 0.14036478529048133, -0...  \n",
              "10597  [-0.07216214826157326, 0.1010708390363707, 0.0...  \n",
              "10598  [-0.07936569826737797, 0.15118979964838652, -0...  \n",
              "10599  [-0.07299363993377773, 0.12335054816609153, 0....  \n",
              "10600  [-0.06328782006464252, 0.12171219619961539, 0....  \n",
              "\n",
              "[10601 rows x 4 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-a3eec2e6-0044-435b-a5c8-78ce3c1075d4\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>masked_address</th>\n",
              "      <th>label</th>\n",
              "      <th>tokens</th>\n",
              "      <th>address_vector</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>id</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>jai balaji distributors new market shop no ** ...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>[jai, balaji, distributors, new, market, shop,...</td>\n",
              "      <td>[-0.06931223549757207, 0.11700991654848282, 0....</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>423</th>\n",
              "      <td>sundram fasteners limited b ** sipcot industri...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>[sundram, fasteners, limited, b, **, sipcot, i...</td>\n",
              "      <td>[-0.0619900122260751, 0.14758910422086188, -0....</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>481</th>\n",
              "      <td>bakul * shardashish township torrent power lim...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>[bakul, shardashish, township, torrent, power,...</td>\n",
              "      <td>[-0.07083080284310861, 0.14458166870855566, -0...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>641</th>\n",
              "      <td>sonkar bhawan near ashok kirana store infront ...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>[sonkar, bhawan, near, ashok, kirana, store, i...</td>\n",
              "      <td>[-0.07039106557285, 0.14437521573040477, -0.00...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4395</th>\n",
              "      <td>jambli maida factory near maharashtra 416121</td>\n",
              "      <td>0.0</td>\n",
              "      <td>[jambli, maida, factory, near, maharashtra, 41...</td>\n",
              "      <td>[-0.06386423349870973, 0.14849421964728765, 0....</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10596</th>\n",
              "      <td>a rajeshwari no *** pampatti street east gate ...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>[a, rajeshwari, no, ***, pampatti, street, eas...</td>\n",
              "      <td>[-0.06445098798100472, 0.14036478529048133, -0...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10597</th>\n",
              "      <td>flat no *** s v srinivasam nehrunagar hyderaba...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>[flat, no, ***, s, v, srinivasam, nehrunagar, ...</td>\n",
              "      <td>[-0.07216214826157326, 0.1010708390363707, 0.0...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10598</th>\n",
              "      <td>madhav nagar katni near post office road murwa...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>[madhav, nagar, katni, near, post, office, roa...</td>\n",
              "      <td>[-0.07936569826737797, 0.15118979964838652, -0...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10599</th>\n",
              "      <td>mig ** sri sai * men s pg road no kphb colony ...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>[mig, **, sri, sai, men, s, pg, road, no, kphb...</td>\n",
              "      <td>[-0.07299363993377773, 0.12335054816609153, 0....</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10600</th>\n",
              "      <td>*th mile tadong near sikkim milk union new hot...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>[th, mile, tadong, near, sikkim, milk, union, ...</td>\n",
              "      <td>[-0.06328782006464252, 0.12171219619961539, 0....</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>10601 rows × 4 columns</p>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-a3eec2e6-0044-435b-a5c8-78ce3c1075d4')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-a3eec2e6-0044-435b-a5c8-78ce3c1075d4 button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-a3eec2e6-0044-435b-a5c8-78ce3c1075d4');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-66e71676-7d90-4fd5-a0a4-b902c26b704d\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-66e71676-7d90-4fd5-a0a4-b902c26b704d')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-66e71676-7d90-4fd5-a0a4-b902c26b704d button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "\n",
              "  <div id=\"id_130a58e5-ab0a-4021-b147-d77fb5e5a050\">\n",
              "    <style>\n",
              "      .colab-df-generate {\n",
              "        background-color: #E8F0FE;\n",
              "        border: none;\n",
              "        border-radius: 50%;\n",
              "        cursor: pointer;\n",
              "        display: none;\n",
              "        fill: #1967D2;\n",
              "        height: 32px;\n",
              "        padding: 0 0 0 0;\n",
              "        width: 32px;\n",
              "      }\n",
              "\n",
              "      .colab-df-generate:hover {\n",
              "        background-color: #E2EBFA;\n",
              "        box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "        fill: #174EA6;\n",
              "      }\n",
              "\n",
              "      [theme=dark] .colab-df-generate {\n",
              "        background-color: #3B4455;\n",
              "        fill: #D2E3FC;\n",
              "      }\n",
              "\n",
              "      [theme=dark] .colab-df-generate:hover {\n",
              "        background-color: #434B5C;\n",
              "        box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "        filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "        fill: #FFFFFF;\n",
              "      }\n",
              "    </style>\n",
              "    <button class=\"colab-df-generate\" onclick=\"generateWithVariable('df')\"\n",
              "            title=\"Generate code using this dataframe.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M7,19H8.4L18.45,9,17,7.55,7,17.6ZM5,21V16.75L18.45,3.32a2,2,0,0,1,2.83,0l1.4,1.43a1.91,1.91,0,0,1,.58,1.4,1.91,1.91,0,0,1-.58,1.4L9.25,21ZM18.45,9,17,7.55Zm-12,3A5.31,5.31,0,0,0,4.9,8.1,5.31,5.31,0,0,0,1,6.5,5.31,5.31,0,0,0,4.9,4.9,5.31,5.31,0,0,0,6.5,1,5.31,5.31,0,0,0,8.1,4.9,5.31,5.31,0,0,0,12,6.5,5.46,5.46,0,0,0,6.5,12Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "    <script>\n",
              "      (() => {\n",
              "      const buttonEl =\n",
              "        document.querySelector('#id_130a58e5-ab0a-4021-b147-d77fb5e5a050 button.colab-df-generate');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      buttonEl.onclick = () => {\n",
              "        google.colab.notebook.generateWithVariable('df');\n",
              "      }\n",
              "      })();\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "    </div>\n",
              "  </div>\n"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "dataframe",
              "variable_name": "df",
              "summary": "{\n  \"name\": \"df\",\n  \"rows\": 10601,\n  \"fields\": [\n    {\n      \"column\": \"id\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 3060,\n        \"min\": 0,\n        \"max\": 10600,\n        \"num_unique_values\": 10601,\n        \"samples\": [\n          2735,\n          2067,\n          1451\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"masked_address\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 10596,\n        \"samples\": [\n          \"s/o $$$$ street no * adesh nagar kotkapura road opposite people petrol pump muktsar punjab 152026\",\n          \"jajpur nahaka road near shiba mandira none cuttack odisha / orissa 755019\",\n          \"no * perumal kovil st poraiyur praiyur melmaruvathoor tamil nadu 603319\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"label\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.511766315719159,\n        \"min\": 0.0,\n        \"max\": 1.0,\n        \"num_unique_values\": 2,\n        \"samples\": [\n          1.0,\n          0.0\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"tokens\",\n      \"properties\": {\n        \"dtype\": \"object\",\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"address_vector\",\n      \"properties\": {\n        \"dtype\": \"object\",\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}"
            }
          },
          "metadata": {},
          "execution_count": 27
        }
      ],
      "source": [
        "df"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [],
      "metadata": {
        "id": "SZgLrRyKfiAP"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "## Additional Features depending on presence of some common keywoeds representing each class\n",
        "# Adding features based on keywords\n",
        "commercial_keywords = r'\\bdistributors\\b|\\bfactory\\b|\\bwarehouse\\b|\\bindustrial\\b|\\bgodown\\b|\\bcomplex\\b|\\bpvt ltd\\b'\n",
        "residential_keywords = r'\\bhouse\\b|\\bflat\\b|\\bstreet\\b|\\bcolony\\b|\\bbungalow\\b'\n",
        "\n",
        "df['is_residential'] = df['masked_address'].str.contains(residential_keywords, case=False, regex=True).astype(int)\n",
        "df['is_commercial'] = df['masked_address'].str.contains(commercial_keywords, case=False, regex=True).astype(int)\n",
        "\n"
      ],
      "metadata": {
        "id": "YsmDABh9SoT1"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "2ZYP0_6iiRFh"
      },
      "outputs": [],
      "source": [
        "## Simpler model logistic regression and as labelled data is less expanding high confidence predictions as training set\n",
        "\n",
        "# import numpy as np\n",
        "# from sklearn.linear_model import LogisticRegression\n",
        "# from sklearn.ensemble import RandomForestClassifier\n",
        "# from sklearn.metrics import classification_report\n",
        "\n",
        "# # Convert lists of vectors to NumPy arrays\n",
        "# X_train_labeled = np.array(df['address_vector'].iloc[:20].tolist())\n",
        "# y_train_labeled = np.array(df['label'].iloc[:20].tolist())\n",
        "# X_unlabeled = np.array(df['address_vector'].iloc[20:].tolist())\n",
        "\n",
        "# # Initial model training\n",
        "# initial_model = LogisticRegression(max_iter=1000)  # Ensure convergence\n",
        "# initial_model.fit(X_train_labeled, y_train_labeled)\n",
        "\n",
        "# # Predict on unlabeled data\n",
        "# y_pred_proba = initial_model.predict_proba(X_unlabeled)\n",
        "# high_confidence_indices = np.where(y_pred_proba.max(axis=1) > 0.5)[0]\n",
        "\n",
        "# # Check if there are any high-confidence predictions\n",
        "# if len(high_confidence_indices) > 0:\n",
        "#     # Select high-confidence predictions\n",
        "#     X_high_confidence = X_unlabeled[high_confidence_indices]\n",
        "#     y_high_confidence = np.argmax(y_pred_proba[high_confidence_indices], axis=1)\n",
        "\n",
        "#     # Combine high-confidence predictions with labeled data\n",
        "#     X_combined = np.vstack((X_train_labeled, X_high_confidence))\n",
        "#     y_combined = np.hstack((y_train_labeled, y_high_confidence))\n",
        "\n",
        "#     # Retrain model\n",
        "#     final_model = RandomForestClassifier()\n",
        "#     final_model.fit(X_combined, y_combined)\n",
        "\n",
        "#     # Evaluate final model on the labeled data\n",
        "#     y_final_pred = final_model.predict(X_train_labeled)\n",
        "#     print(classification_report(y_train_labeled, y_final_pred))\n",
        "# else:\n",
        "#     print(\"No high-confidence predictions found.\")\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "WkI9Kv1kiRFh"
      },
      "outputs": [],
      "source": [
        "## Tried Label Propagarion\n",
        "# from sklearn.semi_supervised import LabelPropagation\n",
        "\n",
        "# # Combine labeled and unlabeled data\n",
        "# X_combined = np.vstack((X_train_labeled, X_unlabeled))\n",
        "# y_combined = np.hstack((y_train_labeled, -1 * np.ones(len(X_unlabeled))))\n",
        "\n",
        "# # Train Label Propagation model\n",
        "# label_prop_model = LabelPropagation(kernel='knn', n_neighbors=10)\n",
        "# label_prop_model.fit(X_combined, y_combined)\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "mxMCwxYaiRFl"
      },
      "outputs": [],
      "source": [
        "# y_pred = label_prop_model.predict(X_unlabeled)\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "iUfZmb8xiRFm"
      },
      "outputs": [],
      "source": [
        "# cnt=0\n",
        "# for i in y_pred:\n",
        "#     if i<=00:\n",
        "#         cnt+=1\n",
        "\n",
        "# cnt"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# # With address_vectors as features trying random forest and as labelled data is less expanding high confidence predictions as training set learning in a self supervised fashion\n",
        "\n",
        "# from sklearn.model_selection import train_test_split\n",
        "# from sklearn.linear_model import LogisticRegression\n",
        "# from sklearn.metrics import accuracy_score, classification_report\n",
        "# from sklearn.preprocessing import StandardScaler\n",
        "\n",
        "\n",
        "\n",
        "# # Step 1: Prepare labeled data\n",
        "# labeled_data = df[df['label'].notna()]\n",
        "# unlabeled_data = df[df['label'].isna()]\n",
        "\n",
        "# X_labeled = np.stack(labeled_data['address_vector'].values)\n",
        "# y_labeled = labeled_data['label']\n",
        "\n",
        "# # Step 2: Split labeled data into train and validation sets\n",
        "# X_train, X_val, y_train, y_val = train_test_split(X_labeled, y_labeled, test_size=0.2, random_state=42)\n",
        "\n",
        "\n",
        "# model = RandomForestClassifier(random_state=42)\n",
        "# model.fit(X_train, y_train)\n",
        "\n",
        "# # Evaluate on validation set\n",
        "# val_predictions = model.predict(X_val)\n",
        "# print(\"Initial model performance:\")\n",
        "# print(classification_report(y_val, val_predictions))\n",
        "\n",
        "# # Step 3: Semi-supervised learning\n",
        "# confidence_threshold = 0.8\n",
        "# X_unlabeled = np.stack(unlabeled_data['address_vector'].values)\n",
        "\n",
        "\n",
        "# unlabeled_predictions = model.predict(X_unlabeled)\n",
        "# unlabeled_probabilities = model.predict_proba(X_unlabeled)\n",
        "\n",
        "# high_confidence_mask = np.max(unlabeled_probabilities, axis=1) > confidence_threshold\n",
        "# high_confidence_predictions = unlabeled_predictions[high_confidence_mask]\n",
        "\n",
        "# # Add high-confidence predictions to the training set\n",
        "# X_train_expanded = np.vstack([X_train, X_unlabeled[high_confidence_mask]])\n",
        "# y_train_expanded = np.concatenate([y_train, high_confidence_predictions])\n",
        "\n",
        "# # Step 5: Retrain model\n",
        "# model_expanded = RandomForestClassifier(random_state=42)\n",
        "# model_expanded.fit(X_train_expanded, y_train_expanded)\n",
        "\n",
        "# # Evaluate expanded model\n",
        "# val_predictions_expanded = model_expanded.predict(X_val)\n",
        "# print(\"\\nExpanded model performance:\")\n",
        "# print(classification_report(y_val, val_predictions_expanded))\n",
        "\n",
        "# # Make predictions on the entire dataset\n",
        "# X_all = np.stack(df['address_vector'].values)\n",
        "# # X_all_scaled = scaler.transform(X_all)\n",
        "# final_predictions = model_expanded.predict(X_all)\n",
        "\n",
        "# # Add predictions to the DataFrame\n",
        "# df['predicted_label'] = final_predictions\n",
        "\n",
        "# df['predicted_label']"
      ],
      "metadata": {
        "id": "QWqZVHQTgBak"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "PAvCJcK4iRFm",
        "outputId": "15113daa-4ab5-4ba8-b689-9cd83c94969d"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "<ipython-input-72-9b8bf94ddd40>:20: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  labeled_data['is_residential'] = labeled_data['is_residential'].astype(int)\n",
            "<ipython-input-72-9b8bf94ddd40>:21: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  labeled_data['is_commercial'] = labeled_data['is_commercial'].astype(int)\n",
            "<ipython-input-72-9b8bf94ddd40>:23: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  unlabeled_data['is_residential'] = unlabeled_data['is_residential'].astype(int)\n",
            "<ipython-input-72-9b8bf94ddd40>:24: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  unlabeled_data['is_commercial'] = unlabeled_data['is_commercial'].astype(int)\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Initial model performance:\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "         0.0       0.67      0.67      0.67         3\n",
            "         1.0       0.50      0.50      0.50         2\n",
            "\n",
            "    accuracy                           0.60         5\n",
            "   macro avg       0.58      0.58      0.58         5\n",
            "weighted avg       0.60      0.60      0.60         5\n",
            "\n",
            "\n",
            "Expanded model performance:\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "         0.0       0.67      0.67      0.67         3\n",
            "         1.0       0.50      0.50      0.50         2\n",
            "\n",
            "    accuracy                           0.60         5\n",
            "   macro avg       0.58      0.58      0.58         5\n",
            "weighted avg       0.60      0.60      0.60         5\n",
            "\n"
          ]
        }
      ],
      "source": [
        "## With address_vectors along with keyword based features defined above\n",
        "# trying random forest and as labelled data is less expanding high confidence predictions as training set learning in a self supervised fashion\n",
        "\n",
        "# 20% of labelled data as test for intial evaluation and tehn predictions as predicted_label\n",
        "\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.ensemble import RandomForestClassifier\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "from sklearn.metrics import classification_report\n",
        "\n",
        "\n",
        "\n",
        "# Step 1: Prepare labeled data\n",
        "labeled_data = df[df['label'].notna()]\n",
        "unlabeled_data = df[df['label'].isna()]\n",
        "\n",
        "# Convert categorical columns to numeric features\n",
        "labeled_data['is_residential'] = labeled_data['is_residential'].astype(int)\n",
        "labeled_data['is_commercial'] = labeled_data['is_commercial'].astype(int)\n",
        "\n",
        "unlabeled_data['is_residential'] = unlabeled_data['is_residential'].astype(int)\n",
        "unlabeled_data['is_commercial'] = unlabeled_data['is_commercial'].astype(int)\n",
        "\n",
        "# Prepare feature matrices\n",
        "X_labeled = np.hstack([\n",
        "    np.stack(labeled_data['address_vector'].values),\n",
        "    labeled_data[['is_residential', 'is_commercial']].values\n",
        "])\n",
        "y_labeled = labeled_data['label']\n",
        "\n",
        "X_unlabeled = np.hstack([\n",
        "    np.stack(unlabeled_data['address_vector'].values),\n",
        "    unlabeled_data[['is_residential', 'is_commercial']].values\n",
        "])\n",
        "\n",
        "# Step 2: Split labeled data into train and validation sets\n",
        "X_train, X_val, y_train, y_val = train_test_split(X_labeled, y_labeled, test_size=0.2, random_state=42)\n",
        "\n",
        "# Initial model\n",
        "model = RandomForestClassifier(random_state=42)\n",
        "model.fit(X_train, y_train)\n",
        "\n",
        "# Evaluate on validation set\n",
        "val_predictions = model.predict(X_val)\n",
        "print(\"Initial model performance:\")\n",
        "print(classification_report(y_val, val_predictions))\n",
        "\n",
        "# Step 3: Semi-supervised learning\n",
        "confidence_threshold = 0.8\n",
        "\n",
        "unlabeled_predictions = model.predict(X_unlabeled)\n",
        "unlabeled_probabilities = model.predict_proba(X_unlabeled)\n",
        "\n",
        "high_confidence_mask = np.max(unlabeled_probabilities, axis=1) > confidence_threshold\n",
        "high_confidence_predictions = unlabeled_predictions[high_confidence_mask]\n",
        "\n",
        "# Add high-confidence predictions to the training set\n",
        "X_train_expanded = np.vstack([X_train, X_unlabeled[high_confidence_mask]])\n",
        "y_train_expanded = np.concatenate([y_train, high_confidence_predictions])\n",
        "\n",
        "# Retrain model\n",
        "model_expanded = RandomForestClassifier(random_state=42,n_estimators=200,max_depth=5)\n",
        "model_expanded.fit(X_train_expanded, y_train_expanded)\n",
        "\n",
        "# Evaluate expanded model\n",
        "val_predictions_expanded = model_expanded.predict(X_val)\n",
        "print(\"\\nExpanded model performance:\")\n",
        "print(classification_report(y_val, val_predictions_expanded))\n",
        "\n",
        "# Make predictions on the entire dataset\n",
        "X_all = np.hstack([\n",
        "    np.stack(df['address_vector'].values),\n",
        "    df[['is_residential', 'is_commercial']].values\n",
        "])\n",
        "final_predictions = model_expanded.predict(X_all)\n",
        "\n",
        "# Add predictions to the DataFrame\n",
        "df['predicted_label_rf'] = final_predictions\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "068aNdWTiRFm",
        "outputId": "fee59d34-4f81-419e-b6ef-13c6b0fb95c8"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Cross-validation scores: [0.6  0.75 0.5  1.   0.5 ]\n",
            "Mean CV score: 0.67\n"
          ]
        }
      ],
      "source": [
        "from sklearn.model_selection import cross_val_score\n",
        "cv_scores = cross_val_score(model, X_labeled, y_labeled, cv=5)\n",
        "\n",
        "print(\"Cross-validation scores:\", cv_scores)\n",
        "print(\"Mean CV score:\", cv_scores.mean())"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(f\"Number of high-confidence samples added: {np.sum(high_confidence_mask)}\")\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "X0egCtL0vW_J",
        "outputId": "dcfbee65-8377-4d4e-864a-2d95c6b4357a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Number of high-confidence samples added: 8481\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#Inference: on the limited labeled data we see that expanding teh training set with high confidence is also not helping betetr the predictions,tried with different threshols of confidence interval at 0.5 we see change\n",
        "# it reflects that there are very lesss high confidence predictions through above method\n",
        "## the predictions by the above model are labelled predicted_label_rf which can be evaluated if the original labels are provided for the rest of the data.\n",
        "df['predicted_label_rf']"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 489
        },
        "id": "5xbKo0s0g23l",
        "outputId": "52899bdd-eb73-4479-e747-2a41dc5a05a8"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "id\n",
              "2        0.0\n",
              "423      1.0\n",
              "481      0.0\n",
              "641      0.0\n",
              "4395     0.0\n",
              "        ... \n",
              "10596    1.0\n",
              "10597    0.0\n",
              "10598    1.0\n",
              "10599    0.0\n",
              "10600    0.0\n",
              "Name: predicted_label_rf, Length: 10601, dtype: float64"
            ],
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>predicted_label_rf</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>id</th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>423</th>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>481</th>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>641</th>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4395</th>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10596</th>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10597</th>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10598</th>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10599</th>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10600</th>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>10601 rows × 1 columns</p>\n",
              "</div><br><label><b>dtype:</b> float64</label>"
            ]
          },
          "metadata": {},
          "execution_count": 37
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "len(X_train_expanded)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "snqOacNU_SK6",
        "outputId": "b23b0799-5668-4475-abd9-f96517252eba"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "3645"
            ]
          },
          "metadata": {},
          "execution_count": 38
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HlhMJ5FFiRFm"
      },
      "source": [
        "## BiLSTM approach\n",
        "\n",
        "The second approach is to use the pre-processed data of the addresess as sequences and use a Bi-LSTM as it has an added advantage to capture both the\n",
        "left and right context in case of text classification. The training is done in a similr fashion of self supervised learning expanding the training set with high confidence predictions."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "OZCkOdeQiRFn"
      },
      "outputs": [],
      "source": [
        "# import numpy as np\n",
        "# from tensorflow.keras.preprocessing.text import Tokenizer\n",
        "# from tensorflow.keras.preprocessing.sequence import pad_sequences\n",
        "# from tensorflow.keras.models import Model\n",
        "# from tensorflow.keras.layers import Input, Embedding, Bidirectional, LSTM, Dense\n",
        "# from tensorflow.keras.optimizers import Adam\n",
        "# from sklearn.model_selection import train_test_split\n",
        "\n",
        "# # Prepare data\n",
        "# tokenizer = Tokenizer()\n",
        "# tokenizer.fit_on_texts(df['masked_address'])\n",
        "# sequences = tokenizer.texts_to_sequences(df['masked_address'])\n",
        "\n",
        "# # Pad sequences\n",
        "# max_length = max(len(seq) for seq in sequences)\n",
        "# padded_sequences = pad_sequences(sequences, maxlen=max_length, padding='post')\n",
        "\n",
        "# # Prepare labeled data\n",
        "# X_labeled = padded_sequences[df['label'].notna()]\n",
        "# y_labeled = df['label'].dropna().values\n",
        "\n",
        "# # Split into train and validation sets\n",
        "# X_train, X_val, y_train, y_val = train_test_split(X_labeled, y_labeled, test_size=0.2, random_state=42)\n",
        "\n",
        "# # Build the model\n",
        "# vocab_size = len(tokenizer.word_index) + 1\n",
        "# embedding_dim = 100\n",
        "\n",
        "# inputs = Input(shape=(max_length,))\n",
        "# x = Embedding(vocab_size, embedding_dim)(inputs)\n",
        "# x = Bidirectional(LSTM(64, return_sequences=True))(x)\n",
        "# x = Bidirectional(LSTM(32))(x)\n",
        "# outputs = Dense(1, activation='sigmoid')(x)\n",
        "\n",
        "# model_lstm= Model(inputs, outputs)\n",
        "# model_lstm.compile(optimizer=Adam(learning_rate=0.001), loss='binary_crossentropy', metrics=['accuracy'])\n",
        "\n",
        "# # Train the model\n",
        "# history = model_lstm.fit(X_train, y_train, validation_data=(X_val, y_val), epochs=10, batch_size=32)\n",
        "\n",
        "# # Evaluate the model\n",
        "# val_loss, val_accuracy = model_lstm.evaluate(X_val, y_val)\n",
        "# print(f\"Validation Accuracy: {val_accuracy}\")\n",
        "\n",
        "# # Make predictions on unlabeled data\n",
        "# X_unlabeled = padded_sequences[df['label'].isna()]\n",
        "# unlabeled_predictions = model_lstm.predict(X_unlabeled)\n",
        "\n",
        "# # Add high-confidence predictions to training set\n",
        "# confidence_threshold = 0.8\n",
        "# high_confidence_mask = np.abs(unlabeled_predictions - 0.5) > (confidence_threshold - 0.5)\n",
        "# high_confidence_predictions = (unlabeled_predictions[high_confidence_mask] > 0.5).astype(int)\n",
        "\n",
        "# X_train_expanded = np.vstack([X_train, X_unlabeled[high_confidence_mask.flatten()]])\n",
        "# y_train_expanded = np.concatenate([y_train, high_confidence_predictions.flatten()])\n",
        "\n",
        "# # Retrain the model with expanded dataset\n",
        "# model_lstm.fit(X_train_expanded, y_train_expanded, epochs=5, batch_size=32)\n",
        "\n",
        "\n",
        "# # Make final predictions\n",
        "# final_predictions = model_lstm.predict(padded_sequences)\n",
        "# df['predicted_label_bilstm'] = (final_predictions > 0.5).astype(int)\n"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "from tensorflow.keras.preprocessing.text import Tokenizer\n",
        "from tensorflow.keras.preprocessing.sequence import pad_sequences\n",
        "from tensorflow.keras.models import Model\n",
        "from tensorflow.keras.layers import Input, Embedding, Bidirectional, LSTM, Dense\n",
        "from tensorflow.keras.optimizers import Adam\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.metrics import classification_report\n",
        "\n",
        "# Prepare data\n",
        "tokenizer = Tokenizer()\n",
        "tokenizer.fit_on_texts(df['masked_address'])\n",
        "sequences = tokenizer.texts_to_sequences(df['masked_address'])\n",
        "\n",
        "# Pad sequences\n",
        "max_length = max(len(seq) for seq in sequences)\n",
        "padded_sequences = pad_sequences(sequences, maxlen=max_length, padding='post')\n",
        "\n",
        "# Prepare labeled data\n",
        "labeled_data = df[df['label'].notna()]\n",
        "unlabeled_data = df[df['label'].isna()]\n",
        "\n",
        "X_labeled = padded_sequences[df['label'].notna()]\n",
        "y_labeled = df['label'].dropna().values\n",
        "\n",
        "# Split into train and validation sets\n",
        "X_train, X_val, y_train, y_val = train_test_split(X_labeled, y_labeled, test_size=0.2, random_state=42)\n",
        "\n",
        "# Build the model\n",
        "vocab_size = len(tokenizer.word_index) + 1\n",
        "embedding_dim = 100\n",
        "\n",
        "inputs = Input(shape=(max_length,))\n",
        "x = Embedding(vocab_size, embedding_dim)(inputs)\n",
        "x = Bidirectional(LSTM(64, return_sequences=True))(x)\n",
        "x = Bidirectional(LSTM(32))(x)\n",
        "outputs = Dense(1, activation='sigmoid')(x)\n",
        "\n",
        "model_lstm = Model(inputs, outputs)\n",
        "model_lstm.compile(optimizer=Adam(learning_rate=0.001), loss='binary_crossentropy', metrics=['accuracy'])\n",
        "\n",
        "# Train the model\n",
        "history = model_lstm.fit(X_train, y_train, validation_data=(X_val, y_val), epochs=10, batch_size=32)\n",
        "\n",
        "# Evaluate the model\n",
        "val_loss, val_accuracy = model_lstm.evaluate(X_val, y_val)\n",
        "print(f\"Validation Accuracy: {val_accuracy}\")\n",
        "\n",
        "# Make predictions on unlabeled data\n",
        "X_unlabeled = padded_sequences[df['label'].isna()]\n",
        "unlabeled_predictions = model_lstm.predict(X_unlabeled)\n",
        "\n",
        "# Convert predictions to labels\n",
        "confidence_threshold = 0.8            # at 0.6 threshold we see some change in X_train_expanded hence low confidence predictions\n",
        "high_confidence_mask = np.abs(unlabeled_predictions - 0.5) > (confidence_threshold - 0.5)\n",
        "high_confidence_predictions = (unlabeled_predictions[high_confidence_mask] > 0.5).astype(int)\n",
        "\n",
        "# Debugging information\n",
        "print(f\"Initial X_train shape: {X_train.shape}\")\n",
        "print(f\"Number of high-confidence samples: {np.sum(high_confidence_mask)}\")\n",
        "print(f\"X_unlabeled shape: {X_unlabeled.shape}\")\n",
        "print(f\"High-confidence X_unlabeled shape: {X_unlabeled[high_confidence_mask.flatten()].shape}\")\n",
        "\n",
        "# Expand training set\n",
        "X_train_expanded = np.vstack([X_train, X_unlabeled[high_confidence_mask.flatten()]])\n",
        "y_train_expanded = np.concatenate([y_train, high_confidence_predictions.flatten()])\n",
        "\n",
        "print(f\"X_train_expanded shape: {X_train_expanded.shape}\")\n",
        "\n",
        "# Retrain the model with expanded dataset\n",
        "model_lstm.fit(X_train_expanded, y_train_expanded, epochs=5, batch_size=32)\n",
        "\n",
        "# Make final predictions\n",
        "final_predictions = model_lstm.predict(padded_sequences)\n",
        "df['predicted_label_bilstm'] = (final_predictions > 0.5).astype(int)\n",
        "\n",
        "# Print the updated classification report\n",
        "print(\"Updated Classification Report:\")\n",
        "print(classification_report(df['label'].dropna(), df['predicted_label_bilstm'][:len(y_labeled)]))\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "hbk0kFmz6zIM",
        "outputId": "71309213-0a0c-4880-e75b-6585628b61be"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/10\n",
            "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 6s/step - accuracy: 0.6250 - loss: 0.6926 - val_accuracy: 0.4000 - val_loss: 0.6945\n",
            "Epoch 2/10\n",
            "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 656ms/step - accuracy: 0.5000 - loss: 0.6883 - val_accuracy: 0.4000 - val_loss: 0.6923\n",
            "Epoch 3/10\n",
            "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 272ms/step - accuracy: 0.7500 - loss: 0.6836 - val_accuracy: 0.8000 - val_loss: 0.6896\n",
            "Epoch 4/10\n",
            "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 156ms/step - accuracy: 0.8750 - loss: 0.6780 - val_accuracy: 0.6000 - val_loss: 0.6869\n",
            "Epoch 5/10\n",
            "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 156ms/step - accuracy: 0.9375 - loss: 0.6710 - val_accuracy: 0.6000 - val_loss: 0.6846\n",
            "Epoch 6/10\n",
            "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 271ms/step - accuracy: 0.9375 - loss: 0.6620 - val_accuracy: 0.6000 - val_loss: 0.6826\n",
            "Epoch 7/10\n",
            "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 170ms/step - accuracy: 0.9375 - loss: 0.6503 - val_accuracy: 0.6000 - val_loss: 0.6807\n",
            "Epoch 8/10\n",
            "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 304ms/step - accuracy: 0.9375 - loss: 0.6353 - val_accuracy: 0.8000 - val_loss: 0.6786\n",
            "Epoch 9/10\n",
            "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 275ms/step - accuracy: 0.9375 - loss: 0.6160 - val_accuracy: 0.8000 - val_loss: 0.6760\n",
            "Epoch 10/10\n",
            "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 146ms/step - accuracy: 0.9375 - loss: 0.5915 - val_accuracy: 0.8000 - val_loss: 0.6728\n",
            "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 65ms/step - accuracy: 0.8000 - loss: 0.6728\n",
            "Validation Accuracy: 0.800000011920929\n",
            "\u001b[1m331/331\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 7ms/step\n",
            "Initial X_train shape: (16, 30)\n",
            "Number of high-confidence samples: 0\n",
            "X_unlabeled shape: (10580, 30)\n",
            "High-confidence X_unlabeled shape: (0, 30)\n",
            "X_train_expanded shape: (16, 30)\n",
            "Epoch 1/5\n",
            "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 35ms/step - accuracy: 0.8750 - loss: 0.5607\n",
            "Epoch 2/5\n",
            "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 58ms/step - accuracy: 0.8750 - loss: 0.5225\n",
            "Epoch 3/5\n",
            "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 57ms/step - accuracy: 0.9375 - loss: 0.4755\n",
            "Epoch 4/5\n",
            "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 33ms/step - accuracy: 0.9375 - loss: 0.4181\n",
            "Epoch 5/5\n",
            "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 58ms/step - accuracy: 0.9375 - loss: 0.3512\n",
            "\u001b[1m332/332\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 4ms/step\n",
            "Updated Classification Report:\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "         0.0       0.82      0.82      0.82        11\n",
            "         1.0       0.80      0.80      0.80        10\n",
            "\n",
            "    accuracy                           0.81        21\n",
            "   macro avg       0.81      0.81      0.81        21\n",
            "weighted avg       0.81      0.81      0.81        21\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "len(X_train_expanded)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1Hbl-Xy36OV8",
        "outputId": "5d81d3e8-10fd-4984-8cae-ea62789411de"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "106"
            ]
          },
          "metadata": {},
          "execution_count": 66
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# # Make predictions on the validation set\n",
        "# val_predictions = model_lstm.predict(X_val)\n",
        "# val_predictions_binary = (val_predictions > 0.5).astype(int)\n",
        "\n",
        "# # Print the classification report\n",
        "# print(\"Validation Classification Report for Bisltm:\")\n",
        "# print(classification_report(y_val, val_predictions_binary))\n"
      ],
      "metadata": {
        "id": "RW1pIFTllnbF"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "CYO9j14FiRFn"
      },
      "outputs": [],
      "source": [
        "## Bislstm approach"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 646
        },
        "id": "_LWv9gToiRFn",
        "outputId": "3ebbdbc5-595f-47e9-a91b-6b2dffe91ebb"
      },
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                                          masked_address  label  \\\n",
              "id                                                                \n",
              "423    sundram fasteners limited b ** sipcot industri...    0.0   \n",
              "0      house padmashali steet jammikunta telangana 50...    1.0   \n",
              "1      s/o $$$$ rly stnchirala pomdl prakasam dt andh...    1.0   \n",
              "3      *** ramnager bagha kheri district uttar prades...    1.0   \n",
              "4        * behind ishvar bhuvan ahmedabad gujarat 380009    1.0   \n",
              "...                                                  ...    ...   \n",
              "10592  rajanna gudem nidamanoor mandal rajannagudam n...    NaN   \n",
              "10593          * hasanpur bithan samastipur bihar 848207    NaN   \n",
              "10594  add rathiala sushil nadone nr iswar resturant ...    NaN   \n",
              "10596  a rajeshwari no *** pampatti street east gate ...    NaN   \n",
              "10597  flat no *** s v srinivasam nehrunagar hyderaba...    NaN   \n",
              "\n",
              "                                                  tokens  \\\n",
              "id                                                         \n",
              "423    [sundram, fasteners, limited, b, **, sipcot, i...   \n",
              "0      [house, padmashali, steet, jammikunta, telanga...   \n",
              "1      [s/o, $$$$, rly, stnchirala, pomdl, prakasam, ...   \n",
              "3      [***, ramnager, bagha, kheri, district, uttar,...   \n",
              "4      [behind, ishvar, bhuvan, ahmedabad, gujarat, 3...   \n",
              "...                                                  ...   \n",
              "10592  [rajanna, gudem, nidamanoor, mandal, rajannagu...   \n",
              "10593      [hasanpur, bithan, samastipur, bihar, 848207]   \n",
              "10594  [add, rathiala, sushil, nadone, nr, iswar, res...   \n",
              "10596  [a, rajeshwari, no, ***, pampatti, street, eas...   \n",
              "10597  [flat, no, ***, s, v, srinivasam, nehrunagar, ...   \n",
              "\n",
              "                                          address_vector  is_residential  \\\n",
              "id                                                                         \n",
              "423    [-0.0619900122260751, 0.14758910422086188, -0....               0   \n",
              "0      [-0.07353603092299928, 0.12748018739069242, -0...               1   \n",
              "1      [-0.08324645092672364, 0.15695248935691938, 0....               0   \n",
              "3      [-0.09386800515711478, 0.15637938498909257, 0....               0   \n",
              "4      [-0.06983228481694634, 0.14309511109197262, 0....               0   \n",
              "...                                                  ...             ...   \n",
              "10592  [-0.07283691660118179, 0.1450538899657947, -0....               0   \n",
              "10593  [-0.06932831750718908, 0.14909402841295596, -0...               0   \n",
              "10594  [-0.09029049869006317, 0.16039965985846735, -0...               0   \n",
              "10596  [-0.06445098798100472, 0.14036478529048133, -0...               1   \n",
              "10597  [-0.07216214826157326, 0.1010708390363707, 0.0...               1   \n",
              "\n",
              "       is_commercial  predicted_label_rf  predicted_label_bilstm  \n",
              "id                                                                \n",
              "423                1                 1.0                       1  \n",
              "0                  0                 1.0                       1  \n",
              "1                  0                 1.0                       1  \n",
              "3                  0                 1.0                       1  \n",
              "4                  0                 1.0                       1  \n",
              "...              ...                 ...                     ...  \n",
              "10592              0                 1.0                       1  \n",
              "10593              0                 0.0                       1  \n",
              "10594              0                 1.0                       1  \n",
              "10596              0                 1.0                       1  \n",
              "10597              0                 0.0                       1  \n",
              "\n",
              "[6540 rows x 8 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-d4affab3-a260-4276-96bb-e73229ebfd63\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>masked_address</th>\n",
              "      <th>label</th>\n",
              "      <th>tokens</th>\n",
              "      <th>address_vector</th>\n",
              "      <th>is_residential</th>\n",
              "      <th>is_commercial</th>\n",
              "      <th>predicted_label_rf</th>\n",
              "      <th>predicted_label_bilstm</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>id</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>423</th>\n",
              "      <td>sundram fasteners limited b ** sipcot industri...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>[sundram, fasteners, limited, b, **, sipcot, i...</td>\n",
              "      <td>[-0.0619900122260751, 0.14758910422086188, -0....</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>house padmashali steet jammikunta telangana 50...</td>\n",
              "      <td>1.0</td>\n",
              "      <td>[house, padmashali, steet, jammikunta, telanga...</td>\n",
              "      <td>[-0.07353603092299928, 0.12748018739069242, -0...</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>s/o $$$$ rly stnchirala pomdl prakasam dt andh...</td>\n",
              "      <td>1.0</td>\n",
              "      <td>[s/o, $$$$, rly, stnchirala, pomdl, prakasam, ...</td>\n",
              "      <td>[-0.08324645092672364, 0.15695248935691938, 0....</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>*** ramnager bagha kheri district uttar prades...</td>\n",
              "      <td>1.0</td>\n",
              "      <td>[***, ramnager, bagha, kheri, district, uttar,...</td>\n",
              "      <td>[-0.09386800515711478, 0.15637938498909257, 0....</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>* behind ishvar bhuvan ahmedabad gujarat 380009</td>\n",
              "      <td>1.0</td>\n",
              "      <td>[behind, ishvar, bhuvan, ahmedabad, gujarat, 3...</td>\n",
              "      <td>[-0.06983228481694634, 0.14309511109197262, 0....</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10592</th>\n",
              "      <td>rajanna gudem nidamanoor mandal rajannagudam n...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>[rajanna, gudem, nidamanoor, mandal, rajannagu...</td>\n",
              "      <td>[-0.07283691660118179, 0.1450538899657947, -0....</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10593</th>\n",
              "      <td>* hasanpur bithan samastipur bihar 848207</td>\n",
              "      <td>NaN</td>\n",
              "      <td>[hasanpur, bithan, samastipur, bihar, 848207]</td>\n",
              "      <td>[-0.06932831750718908, 0.14909402841295596, -0...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10594</th>\n",
              "      <td>add rathiala sushil nadone nr iswar resturant ...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>[add, rathiala, sushil, nadone, nr, iswar, res...</td>\n",
              "      <td>[-0.09029049869006317, 0.16039965985846735, -0...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10596</th>\n",
              "      <td>a rajeshwari no *** pampatti street east gate ...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>[a, rajeshwari, no, ***, pampatti, street, eas...</td>\n",
              "      <td>[-0.06445098798100472, 0.14036478529048133, -0...</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10597</th>\n",
              "      <td>flat no *** s v srinivasam nehrunagar hyderaba...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>[flat, no, ***, s, v, srinivasam, nehrunagar, ...</td>\n",
              "      <td>[-0.07216214826157326, 0.1010708390363707, 0.0...</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>6540 rows × 8 columns</p>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-d4affab3-a260-4276-96bb-e73229ebfd63')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-d4affab3-a260-4276-96bb-e73229ebfd63 button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-d4affab3-a260-4276-96bb-e73229ebfd63');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-ca362b9f-ea5a-40e7-aee0-7b777edc443e\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-ca362b9f-ea5a-40e7-aee0-7b777edc443e')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-ca362b9f-ea5a-40e7-aee0-7b777edc443e button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "\n",
              "    </div>\n",
              "  </div>\n"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "dataframe",
              "summary": "{\n  \"name\": \"df[df['predicted_label_bilstm']==1]\",\n  \"rows\": 6540,\n  \"fields\": [\n    {\n      \"column\": \"id\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 3044,\n        \"min\": 0,\n        \"max\": 10597,\n        \"num_unique_values\": 6540,\n        \"samples\": [\n          4684,\n          396,\n          3519\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"masked_address\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 6540,\n        \"samples\": [\n          \"santipur tiniyali pasali assam 787057\",\n          \"* */* bandlageri mahboobnagar 509001\",\n          \"*** honnekana kelakeri melige near thirthahalli shivamogga karnataka 577415\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"label\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.33333333333333337,\n        \"min\": 0.0,\n        \"max\": 1.0,\n        \"num_unique_values\": 2,\n        \"samples\": [\n          1.0,\n          0.0\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"tokens\",\n      \"properties\": {\n        \"dtype\": \"object\",\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"address_vector\",\n      \"properties\": {\n        \"dtype\": \"object\",\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"is_residential\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0,\n        \"min\": 0,\n        \"max\": 1,\n        \"num_unique_values\": 2,\n        \"samples\": [\n          1,\n          0\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"is_commercial\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0,\n        \"min\": 0,\n        \"max\": 1,\n        \"num_unique_values\": 2,\n        \"samples\": [\n          0,\n          1\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"predicted_label_rf\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.47229846417518534,\n        \"min\": 0.0,\n        \"max\": 1.0,\n        \"num_unique_values\": 2,\n        \"samples\": [\n          0.0,\n          1.0\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"predicted_label_bilstm\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0,\n        \"min\": 1,\n        \"max\": 1,\n        \"num_unique_values\": 1,\n        \"samples\": [\n          1\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}"
            }
          },
          "metadata": {},
          "execution_count": 44
        }
      ],
      "source": [
        "df[df['predicted_label_bilstm']==1]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "b7wDzUe4iRFo"
      },
      "outputs": [],
      "source": [
        "## fine-tuning distilBERT\n",
        "## challenging beacuse the vocabularly is totally different\n",
        "## tried to look around if some open source LLM trained specifically on addresses"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 617
        },
        "id": "reI98C7IiRFo",
        "outputId": "ac6d89e5-5b07-4555-f09d-54f91106b9f7"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Some weights of DistilBertForSequenceClassification were not initialized from the model checkpoint at distilbert-base-uncased and are newly initialized: ['classifier.bias', 'classifier.weight', 'pre_classifier.bias', 'pre_classifier.weight']\n",
            "You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference.\n",
            "/usr/local/lib/python3.10/dist-packages/transformers/training_args.py:1494: FutureWarning: `evaluation_strategy` is deprecated and will be removed in version 4.46 of 🤗 Transformers. Use `eval_strategy` instead\n",
            "  warnings.warn(\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": [
              "\n",
              "    <div>\n",
              "      \n",
              "      <progress value='5' max='5' style='width:300px; height:20px; vertical-align: middle;'></progress>\n",
              "      [5/5 00:03, Epoch 5/5]\n",
              "    </div>\n",
              "    <table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              " <tr style=\"text-align: left;\">\n",
              "      <th>Epoch</th>\n",
              "      <th>Training Loss</th>\n",
              "      <th>Validation Loss</th>\n",
              "      <th>Accuracy</th>\n",
              "      <th>F1</th>\n",
              "      <th>Precision</th>\n",
              "      <th>Recall</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <td>1</td>\n",
              "      <td>No log</td>\n",
              "      <td>0.635499</td>\n",
              "      <td>0.800000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>2</td>\n",
              "      <td>No log</td>\n",
              "      <td>0.635562</td>\n",
              "      <td>0.800000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>3</td>\n",
              "      <td>No log</td>\n",
              "      <td>0.635687</td>\n",
              "      <td>0.800000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>4</td>\n",
              "      <td>No log</td>\n",
              "      <td>0.635879</td>\n",
              "      <td>0.800000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <td>5</td>\n",
              "      <td>No log</td>\n",
              "      <td>0.636134</td>\n",
              "      <td>0.800000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table><p>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/sklearn/metrics/_classification.py:1471: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/metrics/_classification.py:1471: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/metrics/_classification.py:1471: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/metrics/_classification.py:1471: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/metrics/_classification.py:1471: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": []
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/sklearn/metrics/_classification.py:1471: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "{'eval_loss': 0.6361338496208191, 'eval_accuracy': 0.8, 'eval_f1': 0.0, 'eval_precision': 0.0, 'eval_recall': 0.0, 'eval_runtime': 0.0238, 'eval_samples_per_second': 210.275, 'eval_steps_per_second': 42.055, 'epoch': 5.0}\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/sklearn/metrics/_classification.py:1471: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/metrics/_classification.py:1471: UndefinedMetricWarning: Recall and F-score are ill-defined and being set to 0.0 due to no true samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n"
          ]
        }
      ],
      "source": [
        "import torch\n",
        "from transformers import DistilBertTokenizer, DistilBertForSequenceClassification, Trainer, TrainingArguments\n",
        "from sklearn.model_selection import train_test_split\n",
        "from sklearn.metrics import accuracy_score, precision_recall_fscore_support\n",
        "\n",
        "# Check if GPU is available\n",
        "device = torch.device('cuda') if torch.cuda.is_available() else torch.device('cpu')\n",
        "\n",
        "# Load pre-trained tokenizer and model\n",
        "tokenizer = DistilBertTokenizer.from_pretrained('distilbert-base-uncased')\n",
        "model = DistilBertForSequenceClassification.from_pretrained('distilbert-base-uncased', num_labels=2)\n",
        "model.to(device)\n",
        "\n",
        "# Prepare dataset\n",
        "class AddressDataset(torch.utils.data.Dataset):\n",
        "    def __init__(self, encodings, labels):\n",
        "        self.encodings = encodings\n",
        "        self.labels = labels\n",
        "\n",
        "    def __getitem__(self, idx):\n",
        "        item = {key: torch.tensor(val[idx]) for key, val in self.encodings.items()}\n",
        "        item['labels'] = torch.tensor(self.labels[idx])\n",
        "        return item\n",
        "\n",
        "    def __len__(self):\n",
        "        return len(self.labels)\n",
        "\n",
        "# Tokenize and encode the dataset\n",
        "def tokenize_and_encode(texts):\n",
        "    return tokenizer(texts, truncation=True, padding=True, max_length=128)\n",
        "\n",
        "# Prepare labeled data\n",
        "labeled_data = df[df['label'].notna()]\n",
        "train_texts = labeled_data['masked_address'].tolist()\n",
        "train_labels = labeled_data['label'].astype(int).tolist()\n",
        "original_indices = labeled_data.index.tolist()\n",
        "\n",
        "# Split into train and validation sets\n",
        "train_texts, val_texts, train_labels, val_labels,train_indices, val_indices = train_test_split(train_texts, train_labels,original_indices, test_size=0.2,random_state=123)\n",
        "\n",
        "train_encodings = tokenize_and_encode(train_texts)\n",
        "val_encodings = tokenize_and_encode(val_texts)\n",
        "\n",
        "train_dataset = AddressDataset(train_encodings, train_labels)\n",
        "val_dataset = AddressDataset(val_encodings, val_labels)\n",
        "\n",
        "# Define training arguments\n",
        "training_args = TrainingArguments(\n",
        "    output_dir='./results',\n",
        "    num_train_epochs=5,\n",
        "    per_device_train_batch_size=16,\n",
        "    per_device_eval_batch_size=64,\n",
        "    warmup_steps=500,\n",
        "    weight_decay=0.01,\n",
        "    logging_dir='./logs',\n",
        "    logging_steps=10,\n",
        "    evaluation_strategy=\"epoch\",\n",
        ")\n",
        "\n",
        "# Define metrics function\n",
        "def compute_metrics(pred):\n",
        "    labels = pred.label_ids\n",
        "    preds = pred.predictions.argmax(-1)\n",
        "    precision, recall, f1, _ = precision_recall_fscore_support(labels, preds, average='binary')\n",
        "    acc = accuracy_score(labels, preds)\n",
        "    return {\n",
        "        'accuracy': acc,\n",
        "        'f1': f1,\n",
        "        'precision': precision,\n",
        "        'recall': recall\n",
        "    }\n",
        "\n",
        "# Initialize Trainer\n",
        "trainer = Trainer(\n",
        "    model=model,\n",
        "    args=training_args,\n",
        "    train_dataset=train_dataset,\n",
        "    eval_dataset=val_dataset,\n",
        "    compute_metrics=compute_metrics,\n",
        ")\n",
        "\n",
        "# Train the model\n",
        "trainer.train()\n",
        "\n",
        "# Evaluate the model\n",
        "eval_results = trainer.evaluate()\n",
        "print(eval_results)\n",
        "\n",
        "# Make predictions on unlabeled data\n",
        "unlabeled_data = df[df['label'].isna()]\n",
        "unlabeled_texts = unlabeled_data['masked_address'].tolist()\n",
        "unlabeled_encodings = tokenize_and_encode(unlabeled_texts)\n",
        "unlabeled_dataset = AddressDataset(unlabeled_encodings, [0] * len(unlabeled_texts))  # Dummy labels\n",
        "\n",
        "unlabeled_predictions = trainer.predict(unlabeled_dataset)\n",
        "unlabeled_predictions = unlabeled_predictions.predictions.argmax(-1)\n",
        "\n",
        "# Add predictions to the dataframe\n",
        "df.loc[df['label'].isna(), 'predicted_label_ft'] = unlabeled_predictions"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "## pre-processed masked_address\n",
        "##{'eval_loss': 0.6927921175956726, 'eval_accuracy': 0.6, 'eval_f1': 0.0, 'eval_precision': 0.0, 'eval_recall': 0.0, 'eval_runtime': 0.0259, 'eval_samples_per_second': 192.886, 'eval_steps_per_second': 38.577, 'epoch': 3.0}\n",
        "##  accuracy is 0.8 but precision, recall, and F1-score are all 0.0, it indicates that while your model is achieving high accuracy, it is failing to detect the positive class (or is predicting only one class)."
      ],
      "metadata": {
        "id": "l_B4okf-HwoQ"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "F1Mo8gl1IrHQ"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "train_labels"
      ],
      "metadata": {
        "id": "lY3fNHspE57k",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "1bdb4625-5742-482d-b9e4-ecb44d854377"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[1, 0, 1, 1, 1, 1, 0, 0, 1, 0, 1, 0, 0, 1, 0, 1]"
            ]
          },
          "metadata": {},
          "execution_count": 78
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Make predictions on training data\n",
        "train_predictions = trainer.predict(train_dataset)\n",
        "train_pred_labels = train_predictions.predictions.argmax(-1)\n",
        "\n",
        "\n"
      ],
      "metadata": {
        "id": "B2XtnTevBCvP",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 72
        },
        "outputId": "3dc31b54-bb56-4500-bf0d-488cdff5b6df"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": []
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/sklearn/metrics/_classification.py:1471: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# Make predictions on validation data\n",
        "val_predictions = trainer.predict(val_dataset)\n",
        "val_pred_labels = val_predictions.predictions.argmax(-1)\n",
        "\n",
        "\n"
      ],
      "metadata": {
        "id": "60LFBkYSBHhV",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 72
        },
        "outputId": "c0a81ad5-1ffb-4ea1-9b04-4d7effac70b3"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": []
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/sklearn/metrics/_classification.py:1471: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 due to no predicted samples. Use `zero_division` parameter to control this behavior.\n",
            "  _warn_prf(average, modifier, msg_start, len(result))\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [],
      "metadata": {
        "id": "mUjLiHpIE_Oj"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "train_predictions"
      ],
      "metadata": {
        "id": "H8d8HtzxEMgV",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "3d76ab94-dcc0-48cf-8cc7-730708450aa6"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "PredictionOutput(predictions=array([[ 3.31133418e-02, -1.03919394e-01],\n",
              "       [ 2.41089407e-02, -7.97229707e-02],\n",
              "       [ 4.80102152e-02, -8.45208466e-02],\n",
              "       [ 5.08566760e-02, -1.13236167e-01],\n",
              "       [ 3.35088596e-02, -1.17235005e-01],\n",
              "       [ 9.52614118e-06, -9.69187543e-02],\n",
              "       [ 7.77254105e-02, -9.80418697e-02],\n",
              "       [ 4.50487342e-03, -8.76152068e-02],\n",
              "       [ 2.66017187e-02, -1.22171506e-01],\n",
              "       [ 3.57993357e-02, -1.07998312e-01],\n",
              "       [ 2.92233024e-02, -1.24733552e-01],\n",
              "       [ 7.36016929e-02, -8.35067108e-02],\n",
              "       [ 6.17596135e-02, -7.52192885e-02],\n",
              "       [ 5.65124191e-02, -1.06941387e-01],\n",
              "       [ 3.75764742e-02, -7.16967955e-02],\n",
              "       [ 7.46091604e-02, -1.10094219e-01]], dtype=float32), label_ids=array([1, 0, 1, 1, 1, 1, 0, 0, 1, 0, 1, 0, 0, 1, 0, 1]), metrics={'test_loss': 0.7086262702941895, 'test_accuracy': 0.4375, 'test_f1': 0.0, 'test_precision': 0.0, 'test_recall': 0.0, 'test_runtime': 0.0236, 'test_samples_per_second': 679.288, 'test_steps_per_second': 42.455})"
            ]
          },
          "metadata": {},
          "execution_count": 51
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "train_pred_labels"
      ],
      "metadata": {
        "id": "6Z3PmqlMFKdk",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "cfa394f2-f6c4-4726-e0ee-9a5bf91ee3d0"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0])"
            ]
          },
          "metadata": {},
          "execution_count": 52
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "df['predicted_label_ft'][train_indices]=train_pred_labels"
      ],
      "metadata": {
        "id": "bVaPr4LRFNgf",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "cc9ea8d2-e6fc-4f9b-d5aa-dd21cb044550"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "<ipython-input-53-ab4c450627cf>:1: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  df['predicted_label_ft'][train_indices]=train_pred_labels\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "df['predicted_label_ft'][val_indices]=val_pred_labels"
      ],
      "metadata": {
        "id": "XqwgGrpxFi6_",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "600fccab-ee43-4ce2-8cd5-b1b4aab07606"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "<ipython-input-54-add39f82565b>:1: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  df['predicted_label_ft'][val_indices]=val_pred_labels\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "df"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 837
        },
        "id": "tuyCL-ksDllb",
        "outputId": "fd480d4f-f925-46cc-cb5a-6e4880f4233c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "                                          masked_address  label  \\\n",
              "id                                                                \n",
              "2      jai balaji distributors new market shop no ** ...    0.0   \n",
              "423    sundram fasteners limited b ** sipcot industri...    0.0   \n",
              "481    bakul * shardashish township torrent power lim...    0.0   \n",
              "641    sonkar bhawan near ashok kirana store infront ...    0.0   \n",
              "4395        jambli maida factory near maharashtra 416121    0.0   \n",
              "...                                                  ...    ...   \n",
              "10596  a rajeshwari no *** pampatti street east gate ...    NaN   \n",
              "10597  flat no *** s v srinivasam nehrunagar hyderaba...    NaN   \n",
              "10598  madhav nagar katni near post office road murwa...    NaN   \n",
              "10599  mig ** sri sai * men s pg road no kphb colony ...    NaN   \n",
              "10600  *th mile tadong near sikkim milk union new hot...    NaN   \n",
              "\n",
              "                                                  tokens  \\\n",
              "id                                                         \n",
              "2      [jai, balaji, distributors, new, market, shop,...   \n",
              "423    [sundram, fasteners, limited, b, **, sipcot, i...   \n",
              "481    [bakul, shardashish, township, torrent, power,...   \n",
              "641    [sonkar, bhawan, near, ashok, kirana, store, i...   \n",
              "4395   [jambli, maida, factory, near, maharashtra, 41...   \n",
              "...                                                  ...   \n",
              "10596  [a, rajeshwari, no, ***, pampatti, street, eas...   \n",
              "10597  [flat, no, ***, s, v, srinivasam, nehrunagar, ...   \n",
              "10598  [madhav, nagar, katni, near, post, office, roa...   \n",
              "10599  [mig, **, sri, sai, men, s, pg, road, no, kphb...   \n",
              "10600  [th, mile, tadong, near, sikkim, milk, union, ...   \n",
              "\n",
              "                                          address_vector  is_residential  \\\n",
              "id                                                                         \n",
              "2      [-0.06931223549757207, 0.11700991654848282, 0....               0   \n",
              "423    [-0.0619900122260751, 0.14758910422086188, -0....               0   \n",
              "481    [-0.07083080284310861, 0.14458166870855566, -0...               0   \n",
              "641    [-0.07039106557285, 0.14437521573040477, -0.00...               0   \n",
              "4395   [-0.06386423349870973, 0.14849421964728765, 0....               0   \n",
              "...                                                  ...             ...   \n",
              "10596  [-0.06445098798100472, 0.14036478529048133, -0...               1   \n",
              "10597  [-0.07216214826157326, 0.1010708390363707, 0.0...               1   \n",
              "10598  [-0.07936569826737797, 0.15118979964838652, -0...               0   \n",
              "10599  [-0.07299363993377773, 0.12335054816609153, 0....               1   \n",
              "10600  [-0.06328782006464252, 0.12171219619961539, 0....               0   \n",
              "\n",
              "       is_commercial  predicted_label_rf  predicted_label_bilstm  \\\n",
              "id                                                                 \n",
              "2                  1                 0.0                       0   \n",
              "423                1                 1.0                       1   \n",
              "481                0                 0.0                       0   \n",
              "641                0                 0.0                       0   \n",
              "4395               1                 0.0                       1   \n",
              "...              ...                 ...                     ...   \n",
              "10596              0                 1.0                       1   \n",
              "10597              0                 0.0                       1   \n",
              "10598              0                 1.0                       0   \n",
              "10599              0                 0.0                       0   \n",
              "10600              0                 0.0                       1   \n",
              "\n",
              "       predicted_label_ft  \n",
              "id                         \n",
              "2                     0.0  \n",
              "423                   0.0  \n",
              "481                   0.0  \n",
              "641                   0.0  \n",
              "4395                  0.0  \n",
              "...                   ...  \n",
              "10596                 0.0  \n",
              "10597                 0.0  \n",
              "10598                 0.0  \n",
              "10599                 0.0  \n",
              "10600                 0.0  \n",
              "\n",
              "[10601 rows x 9 columns]"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-cc5c33fe-9410-4a78-b387-e6183286644c\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>masked_address</th>\n",
              "      <th>label</th>\n",
              "      <th>tokens</th>\n",
              "      <th>address_vector</th>\n",
              "      <th>is_residential</th>\n",
              "      <th>is_commercial</th>\n",
              "      <th>predicted_label_rf</th>\n",
              "      <th>predicted_label_bilstm</th>\n",
              "      <th>predicted_label_ft</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>id</th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>jai balaji distributors new market shop no ** ...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>[jai, balaji, distributors, new, market, shop,...</td>\n",
              "      <td>[-0.06931223549757207, 0.11700991654848282, 0....</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>423</th>\n",
              "      <td>sundram fasteners limited b ** sipcot industri...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>[sundram, fasteners, limited, b, **, sipcot, i...</td>\n",
              "      <td>[-0.0619900122260751, 0.14758910422086188, -0....</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>481</th>\n",
              "      <td>bakul * shardashish township torrent power lim...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>[bakul, shardashish, township, torrent, power,...</td>\n",
              "      <td>[-0.07083080284310861, 0.14458166870855566, -0...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>641</th>\n",
              "      <td>sonkar bhawan near ashok kirana store infront ...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>[sonkar, bhawan, near, ashok, kirana, store, i...</td>\n",
              "      <td>[-0.07039106557285, 0.14437521573040477, -0.00...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4395</th>\n",
              "      <td>jambli maida factory near maharashtra 416121</td>\n",
              "      <td>0.0</td>\n",
              "      <td>[jambli, maida, factory, near, maharashtra, 41...</td>\n",
              "      <td>[-0.06386423349870973, 0.14849421964728765, 0....</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10596</th>\n",
              "      <td>a rajeshwari no *** pampatti street east gate ...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>[a, rajeshwari, no, ***, pampatti, street, eas...</td>\n",
              "      <td>[-0.06445098798100472, 0.14036478529048133, -0...</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10597</th>\n",
              "      <td>flat no *** s v srinivasam nehrunagar hyderaba...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>[flat, no, ***, s, v, srinivasam, nehrunagar, ...</td>\n",
              "      <td>[-0.07216214826157326, 0.1010708390363707, 0.0...</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10598</th>\n",
              "      <td>madhav nagar katni near post office road murwa...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>[madhav, nagar, katni, near, post, office, roa...</td>\n",
              "      <td>[-0.07936569826737797, 0.15118979964838652, -0...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1.0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10599</th>\n",
              "      <td>mig ** sri sai * men s pg road no kphb colony ...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>[mig, **, sri, sai, men, s, pg, road, no, kphb...</td>\n",
              "      <td>[-0.07299363993377773, 0.12335054816609153, 0....</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>10600</th>\n",
              "      <td>*th mile tadong near sikkim milk union new hot...</td>\n",
              "      <td>NaN</td>\n",
              "      <td>[th, mile, tadong, near, sikkim, milk, union, ...</td>\n",
              "      <td>[-0.06328782006464252, 0.12171219619961539, 0....</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>1</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>10601 rows × 9 columns</p>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-cc5c33fe-9410-4a78-b387-e6183286644c')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-cc5c33fe-9410-4a78-b387-e6183286644c button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-cc5c33fe-9410-4a78-b387-e6183286644c');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-fe6e8f8e-4108-46d6-aac1-6bf17e484c88\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-fe6e8f8e-4108-46d6-aac1-6bf17e484c88')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-fe6e8f8e-4108-46d6-aac1-6bf17e484c88 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "\n",
              "  <div id=\"id_a63aedef-d21e-4f51-925d-b3e2fce589e5\">\n",
              "    <style>\n",
              "      .colab-df-generate {\n",
              "        background-color: #E8F0FE;\n",
              "        border: none;\n",
              "        border-radius: 50%;\n",
              "        cursor: pointer;\n",
              "        display: none;\n",
              "        fill: #1967D2;\n",
              "        height: 32px;\n",
              "        padding: 0 0 0 0;\n",
              "        width: 32px;\n",
              "      }\n",
              "\n",
              "      .colab-df-generate:hover {\n",
              "        background-color: #E2EBFA;\n",
              "        box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "        fill: #174EA6;\n",
              "      }\n",
              "\n",
              "      [theme=dark] .colab-df-generate {\n",
              "        background-color: #3B4455;\n",
              "        fill: #D2E3FC;\n",
              "      }\n",
              "\n",
              "      [theme=dark] .colab-df-generate:hover {\n",
              "        background-color: #434B5C;\n",
              "        box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "        filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "        fill: #FFFFFF;\n",
              "      }\n",
              "    </style>\n",
              "    <button class=\"colab-df-generate\" onclick=\"generateWithVariable('df')\"\n",
              "            title=\"Generate code using this dataframe.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "       width=\"24px\">\n",
              "    <path d=\"M7,19H8.4L18.45,9,17,7.55,7,17.6ZM5,21V16.75L18.45,3.32a2,2,0,0,1,2.83,0l1.4,1.43a1.91,1.91,0,0,1,.58,1.4,1.91,1.91,0,0,1-.58,1.4L9.25,21ZM18.45,9,17,7.55Zm-12,3A5.31,5.31,0,0,0,4.9,8.1,5.31,5.31,0,0,0,1,6.5,5.31,5.31,0,0,0,4.9,4.9,5.31,5.31,0,0,0,6.5,1,5.31,5.31,0,0,0,8.1,4.9,5.31,5.31,0,0,0,12,6.5,5.46,5.46,0,0,0,6.5,12Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "    <script>\n",
              "      (() => {\n",
              "      const buttonEl =\n",
              "        document.querySelector('#id_a63aedef-d21e-4f51-925d-b3e2fce589e5 button.colab-df-generate');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      buttonEl.onclick = () => {\n",
              "        google.colab.notebook.generateWithVariable('df');\n",
              "      }\n",
              "      })();\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "    </div>\n",
              "  </div>\n"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "dataframe",
              "variable_name": "df",
              "summary": "{\n  \"name\": \"df\",\n  \"rows\": 10601,\n  \"fields\": [\n    {\n      \"column\": \"id\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 3060,\n        \"min\": 0,\n        \"max\": 10600,\n        \"num_unique_values\": 10601,\n        \"samples\": [\n          2735,\n          2067,\n          1451\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"masked_address\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 10596,\n        \"samples\": [\n          \"s/o $$$$ street no * adesh nagar kotkapura road opposite people petrol pump muktsar punjab 152026\",\n          \"jajpur nahaka road near shiba mandira none cuttack odisha / orissa 755019\",\n          \"no * perumal kovil st poraiyur praiyur melmaruvathoor tamil nadu 603319\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"label\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.511766315719159,\n        \"min\": 0.0,\n        \"max\": 1.0,\n        \"num_unique_values\": 2,\n        \"samples\": [\n          1.0,\n          0.0\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"tokens\",\n      \"properties\": {\n        \"dtype\": \"object\",\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"address_vector\",\n      \"properties\": {\n        \"dtype\": \"object\",\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"is_residential\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0,\n        \"min\": 0,\n        \"max\": 1,\n        \"num_unique_values\": 2,\n        \"samples\": [\n          1,\n          0\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"is_commercial\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0,\n        \"min\": 0,\n        \"max\": 1,\n        \"num_unique_values\": 2,\n        \"samples\": [\n          0,\n          1\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"predicted_label_rf\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.4336170381175383,\n        \"min\": 0.0,\n        \"max\": 1.0,\n        \"num_unique_values\": 2,\n        \"samples\": [\n          1.0,\n          0.0\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"predicted_label_bilstm\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0,\n        \"min\": 0,\n        \"max\": 1,\n        \"num_unique_values\": 2,\n        \"samples\": [\n          1,\n          0\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"predicted_label_ft\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 0.0,\n        \"min\": 0.0,\n        \"max\": 0.0,\n        \"num_unique_values\": 1,\n        \"samples\": [\n          0.0\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}"
            }
          },
          "metadata": {},
          "execution_count": 74
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#df.to_csv('predictions_rf_bilstm_distillbert_refined_data.csv')"
      ],
      "metadata": {
        "id": "VBm0D0LJF0W_"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# df['predicted_label_ft'].value_counts()"
      ],
      "metadata": {
        "id": "TspZoFIPF9a6"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "## Tried many iterations of confidence thresholds with the 21 labeled data and preprocessed masked address but for rf couldn't find points beyound confidence interval toa dd to train expanded same with bislstm, the training data is too less"
      ],
      "metadata": {
        "id": "hfTW149xRloP"
      },
      "execution_count": null,
      "outputs": []
    }
  ],
  "metadata": {
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.10.8"
    },
    "colab": {
      "provenance": [],
      "gpuType": "T4"
    },
    "accelerator": "GPU"
  },
  "nbformat": 4,
  "nbformat_minor": 0
}