Basic Image Classification with TensorFlow

This project demonstrates a basic image classification task using TensorFlow. The goal is to classify handwritten digits from the MNIST dataset.

Setup and Requirements

• Python 3.x
• TensorFlow
• Matplotlib
• NumPy

Code Explanation

Task 1: Introduction

• Import TensorFlow library and set verbosity.
• Check the TensorFlow version.

Task 2: The Dataset

• Import the MNIST dataset.
• Display the shapes of the imported arrays.
• Plot an example image from the dataset.

Task 3: One Hot Encoding

• Encode the labels using one-hot encoding.

Task 5: Preprocessing the Examples

• Reshape the input images.
• Normalize the pixel values.

Task 6: Creating a Model

• Define and compile the neural network model.

Task 7: Training the Model

• Train the model using the training data.

Task 8: Evaluating the Model

• Evaluate the model's performance on the test set.

Task 9: Predictions

• Generate predictions on the test set.
• Visualize the predictions using Matplotlib.

Results

• Trained the model with an accuracy of 96% on the test set.
• Visualized predictions on a subset of the test set using Matplotlib.

Usage

1. Clone the repository.
2. Install the required dependencies.
3. Run the main script to train the model and evaluate its performance.

Acknowledgements

• Special thanks to the creators of the MNIST dataset.
• This project is licensed under the MIT License