Skip to content

Latest commit

 

History

History
18 lines (13 loc) · 1.22 KB

README.md

File metadata and controls

18 lines (13 loc) · 1.22 KB

neural-net

This repository implements a neural network using 2d arrays and math in pure java. No other frameworks and libraries used.

mock_network.webp

You've come to the right place if :

  • you're new to ML
  • fed up with those heavily abstracted libraries like scikit learn, tensorflow, numpy (yes i'll consider numpy as abstracted!)
  • want to understand how these math formulas are translated to code
  • aaaand want to see it all in java (cause why not)

There are a series of medium blogs which take you through the implementation.

  1. Defines the goals and the base object - Matrix
  2. Explains how feedforward is implemented
  3. Explains backpropogation
  4. Tests the implementation on the MNIST dataset