House Price Prediction Project

This project is a House Price Prediction Application developed for a final year project. It uses machine learning to predict house prices based on user input. The application has a React.js frontend and a Flask backend, which uses a pre-trained machine learning model.

---

Features

• Predict house prices based on input features such as bedrooms, bathrooms, and house condition.
• Modern and responsive user interface built with React.js.
• Scalable Flask API that serves predictions from a pre-trained model.
• Easily deployable on localhost for testing and demonstrations.

---

Project Structure

Backend

• model/: Contains the trained machine learning model (house_price_model.pkl) and the scaler (scaler.pkl).
• notebook/: Jupyter notebooks used for training the machine learning model.
• app.py: Flask application that serves as the API.
• requirements.txt: List of dependencies needed for the backend.

Frontend

• public/: Public assets for the React application.
• src/: Source files for the React application.
  • App.js: Main component handling the UI.
  • components/: Reusable components.
• package.json: Configuration file listing dependencies and scripts for the frontend.

---

How to Run the Project

Step 1: Clone the Repository

Make sure you have Git installed on your system. Open your terminal and run:

git clone https://github.com/Rishiraj8/house_prediction.git
cd House

---

Step 2: Set Up the Backend

1. Navigate to the backend folder:

   cd backend

2. Create and activate a Python virtual environment (optional but recommended):

   python -m venv venv
   source venv/bin/activate  # On Windows, use venv\Scripts\activate

3. Install the required Python dependencies:

   pip install -r requirements.txt

4. Start the Flask server:

   python app.py

   • The Flask server will now run on http://127.0.0.1:5000 (localhost).
   • Keep this terminal running.

---

Step 3: Set Up the Frontend

1. Open a new terminal and navigate to the frontend folder:

   cd frontend

2. Install the required Node.js dependencies:

   npm install

3. Start the React development server:

   npm run dev

   • The React app will now run on http://127.0.0.1:5173 (localhost).
   • Keep this terminal running.

---

Step 4: Use the Application

1. Open your browser and go to http://127.0.0.1:5173.
2. Enter the required details in the form (e.g., number of bedrooms, bathrooms, etc.).
3. Click the Predict Price button.
4. The predicted house price will be displayed on the screen.

---

Notes

• Ensure both the Flask backend and React frontend are running simultaneously.
• If you encounter issues, check the terminal logs for errors.
• You can modify the machine learning model or retrain it using the Jupyter notebook in the notebook/ folder.

---

Model Description

1. Dataset:

   • The model was trained using a dataset (house_data.csv) containing features related to house properties and their respective prices.

2. Features Used:

   • Number of bedrooms
   • Number of bathrooms
   • Living area (in square feet)
   • Condition of the house (scaled from 1 to 5)
   • Number of schools nearby

3. Target Variable:

   • Price: The dependent variable representing the house price.

4. Data Preprocessing:

   • The feature variables were scaled using StandardScaler to normalize the data. This ensures that features with varying magnitudes do not disproportionately influence the model.

5. Model Used:

   • Random Forest Regressor: A machine learning ensemble model that uses multiple decision trees to predict the target variable.
     • Key Parameters:
       • n_estimators=100: The model uses 100 decision trees for making predictions.
       • random_state=42: Ensures reproducibility of results.
     • Why Random Forest:
       • Handles non-linear relationships well.
       • Resistant to overfitting when tuned properly.
       • Capable of handling feature importance, making it suitable for regression tasks like house price prediction.

6. Model Evaluation:

   • R² Score (Coefficient of Determination):
     • Training R² Score: Measures how well the model fits the training data.
     • Testing R² Score: Indicates the model's predictive performance on unseen data.

7. Model Saving:

   • The trained model (house_price_model.pkl) and the scaler (scaler.pkl) were serialized and saved for future use. These files are located in the backend\model directory.

---

Acknowledgments

This project was developed as a final year project for a school friend of mine.