Neural Network Exploration (MNIST)

This project explores the basics of neural networks using Keras and TensorFlow.
A simple fully connected neural network is trained on the MNIST handwritten digits dataset to perform digit classification.

Overview

• Loads and preprocesses the MNIST dataset
• Builds a feedforward neural network
• Trains the model using stochastic gradient descent
• Predicts and visualizes results on test images

Model Architecture

• Input: 28 × 28 grayscale images (flattened)
• Hidden Layer 1: 300 neurons, ReLU activation
• Hidden Layer 2: 100 neurons, ReLU activation
• Output Layer: 10 neurons, Softmax activation

Technologies Used

• Python
• TensorFlow / Keras
• NumPy
• Matplotlib