Skip to content

Latest commit

 

History

History
48 lines (31 loc) · 1.27 KB

README.md

File metadata and controls

48 lines (31 loc) · 1.27 KB
Raw

Neural Network from Scratch

This project implements a basic neural network from scratch using Python. The network can be trained to classify or predict based on input data without the use of external machine learning libraries.

Features

  • Forward propagation
  • Backpropagation
  • Adjustable learning rate
  • Multi-layer architecture

How to Run

  1. Clone the repository:
git clone https://github.com/darwin-luque/neural-network.git
cd neural-network
  1. Install dependencies:
pip install -r requirements.txt
  1. Run the script:
python testing.py

Note: Generate your own dataset. Maybe try using the XOR problem to test the network.

How It Works

  • Forward Propagation: The network computes the output by passing inputs through multiple layers.
  • Backpropagation: The error is computed and propagated back to update weights.

Example

To showcase how the neural network works, you can run it on a simple dataset (e.g., XOR problem) and evaluate the model’s accuracy.

Future Improvements

  • Adding more complex activation functions like ReLU or Tanh.
  • Implementing optimization techniques like momentum or learning rate decay.
  • Expanding the network to include convolutional neural networks (CNNs) or recurrent neural networks (RNNs).