https://github.com/sverrenystad/neural-network
A simple implementation of a modifiable Multi Layer Perceptron in Python
https://github.com/sverrenystad/neural-network
deep-learning machine-learning multilayer-perceptron neural-network
Last synced: 16 days ago
JSON representation
A simple implementation of a modifiable Multi Layer Perceptron in Python
- Host: GitHub
- URL: https://github.com/sverrenystad/neural-network
- Owner: SverreNystad
- Created: 2024-03-04T22:24:10.000Z (over 1 year ago)
- Default Branch: main
- Last Pushed: 2024-06-03T15:51:40.000Z (over 1 year ago)
- Last Synced: 2025-09-20T12:54:39.451Z (19 days ago)
- Topics: deep-learning, machine-learning, multilayer-perceptron, neural-network
- Language: Jupyter Notebook
- Homepage:
- Size: 617 KB
- Stars: 1
- Watchers: 1
- Forks: 0
- Open Issues: 6
-
Metadata Files:
- Readme: readme.md
Awesome Lists containing this project
README
# Multilayer Perceptron (Aka Feedforward Neural Network)
This project is a simple implementation of a neural network in Python. It utilizes a sigmoid activation function and mean squared error loss to train the model on a synthetic dataset. The network is built from scratch, providing insights into the basics of neural network operations including forward and backward propagation.
## Features
- Custom neural network implementation.
- Many activation functions like Sigmoid and ReLU and loss functions like mean squared error.
- Training and evaluation on a synthetic dataset.
- Visualization of training loss and neuron weights.
- Exploratory data analysis of the synthetic dataset. [EDA](src/eda.ipynb)
## Setup
Ensure you have Python installed, then set up your environment:
```bash
pip install -r requirements.txt
```
## Usage
Execute the program to see the Neural Network in action:
```bash
python main.py
```
## Testing
To run the tests, execute:
```bash
pytest
```