A simple implementation of a Neural Network from scratch in pure Python — without using deep learning libraries.
This project is aimed at building intuition about how neural networks work under the hood, covering forward propagation, backpropagation, and training on toy datasets.
- Implemented using only Python + NumPy
- Supports:
- Forward propagation
- Backpropagation with gradient descent
- Multiple hidden layers
- Simple training loop on sample data
- Clear, educational code with comments
- Initialize the network with chosen input, hidden, and output sizes.
- Train the network using gradient descent.
- Observe loss decreasing across epochs.
For more detail understanding you may visit This Youtube Playlist