Skip to content

Latest commit

 

History

History
18 lines (11 loc) · 803 Bytes

File metadata and controls

18 lines (11 loc) · 803 Bytes

Simple Neural Network in Python

This project was something I started and later used for both my AI and Machine Learning courses, with additions and alterations for each. The program utilizes NumPy for computations, and PyPlot for a couple simple visualizations. The program demo uses sklearn's Digits dataset.

File overview

  • Transform.py: a short script that prepares the Digits dataset for use in the neural network

  • NeuralNet.py: the neural network class

  • Demo.py: a script that demonstrates the neural network

  • digits.csv: the Digits dataset saved from sklearn

Build instructions

First run Transform.py to transform the data into something more suitable, and run with Demo.py, which also contains hyperparameters to tinker with.