Fashion MNIST Neural Network with TensorFlow and Keras

This project demonstrates how to build, train, and evaluate a fully connected neural network (FCN) for image classification using the Fashion MNIST dataset. The project uses TensorFlow and Keras for deep learning tasks and Matplotlib for visualization.

Dataset

The dataset used in this project is the Fashion MNIST dataset, which consists of 60,000 training images and 10,000 testing images of 28x28 grayscale images across 10 different categories. Each image corresponds to a clothing item.

Project Structure

1. Load Data:

   • The Fashion MNIST dataset is loaded directly from TensorFlow's built-in datasets.
   • The dataset is split into training and test sets.

2. Data Preprocessing:

   • The images are reshaped from 28x28 matrices into 1D vectors (flattened to 784 pixels).
   • Pixel values are normalized to be in the range of 0 to 1.
   • Labels are converted to one-hot encoding for classification.

3. Neural Network Architecture:

   • Input layer: 784 nodes (one for each pixel).
   • Hidden layer 1: 200 neurons, ReLU activation.
   • Hidden layer 2: 100 neurons, ReLU activation.
   • Output layer: 10 neurons, softmax activation (for 10 clothing categories).

4. Model Compilation:

   • Loss function: Categorical Cross-Entropy.
   • Optimizer: Stochastic Gradient Descent (SGD).
   • Metric: Accuracy.

5. Training:

   • The model is trained for 10 epochs with a batch size of 5.
   • Training time is recorded.

6. Evaluation:

   • Test accuracy is calculated.
   • Random samples from the test dataset are visualized with their predicted labels.

Requirements

To run this project, you need the following dependencies:

• TensorFlow
• NumPy
• Matplotlib
• Scikit-learn

You can install the required dependencies using pip:

pip install tensorflow numpy matplotlib scikit-learn