ML From Scratch

This repository contains implementations of classical machine learning algorithms and basic neural networks written from scratch for educational purposes.

Implemented models

• k-Nearest Neighbor (kNN)
• Softmax Linear Classifier
• Two-Layer Neural Network (Fully Connected)

Repository structure

• src/ — reusable ML implementations
• demos/ — notebooks demonstrating how to use the models
• data/ — datasets (ignored by git)
• artifacts/ — saved models and outputs (ignored by git)