This project implements a handwritten digit recognition system using the MNIST dataset and a Neural Network built with TensorFlow and Keras Api.
The MNIST dataset contains 70,000 grayscale images of handwritten digits (0 to 9), each sized 28×28 pixels. In this project, we use a fully connected feedforward neural network (Multi-Layer Perceptron) to learn and classify digits.
The model is trained on 60,000 images and evaluated on 10,000 test images, demonstrating key concepts in:
- Data preprocessing (reshaping, normalization)
- Neural network architecture design
- Model training using backpropagation and gradient descent
- Evaluation metrics (accuracy, loss curves)
- Making predictions on new handwritten digits
This project is a practical introduction to deep learning for image classification tasks.