Neural Network from Scratch

This repository contains some Machine Learning / Deep Learning Algorithms implemented from scratch. The purpose of the implementations is to be easy to understand and are not optimized for performance.

Dependencies

You can install the necessary dependencies with uv or pip:

uv install

pip install -r requirements.txt

Main Dependencies

numpy
tensorflow # for test datasets

Simple Neural Network

A simple neural network implementation using Stochastic Gradient Descent (SGD).

• Code
• Explanation

Neural Network using Reverse Mode Auto-Diff

Implementation of a neural network using Reverse Model Automatic Differentiation Stochastic Gradient Descent.

• To Be implemented