ANN-Number-Recognition

A simple artificial neural network (ANN) implementation for recognizing handwritten digits using the MNIST dataset.

Project Structure

ANN-Number-Recognition/
├── model.py             # Neural network model implementation
├── utils.py            # Helper functions
├── train.py            # Training script
├── test.py             # Testing and visualization
├── requirements.txt    # Python dependencies
└── README.md           # Project documentation

Installation

pip install -r requirements.txt

Usage

Training the network

python train.py

This will train the model on the MNIST dataset and save the parameters in a file defaultly named as mnist.pth.

Testing / Inference

python test.py

This will evaluate the model on test samples and visualize predictions.

Acknowledgements

• MNIST dataset from Yann LeCun's website

License

This project is licensed under the MIT License.