Back to browse
GitHub Repository

Companion code for Machine Learning From Scratch — 10 core ML algorithms built from scratch with NumPy, compared with Scikit-learn and PyTorch.

11 starsJupyter Notebook

I built 10 ML algos from scratch because fit() predict() are not enough

by akmoleksandr·Jun 9, 2026·4 points·1 comment

AI Analysis

MidCozyNiche Gem

Yet another ML-from-scratch teaching repo in a crowded educational space.

Strengths
  • Five-stage framework (intuition, formalization, implementation, test, tips) structures learning well.
  • Direct comparison with Scikit-learn and PyTorch validates from-scratch implementations.
Weaknesses
  • Reference implementation for a book, not a novel tool or technique.
  • Building ML algos from scratch is well-trodden educational territory.
Category
Target Audience

Students learning ML fundamentals, self-taught developers

Similar To

grokking-deep-learning · ML-From-Scratch · fastai

Similar Projects

AI/ML●●Solid

ApplyPilot – Self-Hosted AI-Powered Job Search Companion. Open Source

Five-agent LangGraph pipeline beats single-prompt tools like Simplify and Teal.

Big BrainSolve My Problem
eliornl
101mo ago