60+ ML/AI/DL/RL/Bayesian algorithms implemented from scratch in NumPy -- plus mlscratch, a pip-installable package (pip install scratchkit) with a consistent, scikit-learn-style API and 1,100+ tests.
Interview-ready ML algorithms repository covering 30+ algorithms from scratch using NumPy -- supervised, unsupervised, ensemble, deep learning, and reinforcement learning. Includes theory, math, implementation, and complexity analysis
Applied ML and statistics portfolio: causal inference (regression discontinuity), classical ML pipelines, computer vision, NLP, and a tested NumPy-only ml_from_scratch package. 39/39 notebooks execute, including a 20-notebook retrospective of the 2019 work.
An interactive Heart Disease Prediction web app. The core machine learning models (Random Forest, Logistic Regression, KNN) are built entirely from scratch using mathematical principles in NumPy, without relying on scikit-learn.
My Machine Learning portfolio – practicing and implementing algorithms from scratch as I learn.
A comprehensive, hand-coded survival guide to Artificial Intelligence and Machine Learning fundamentals. Features step-by-step mathematical derivations, algorithmic implementations from scratch (A* Search, Genetic Algorithms, Minimax, Naive Bayes, Gradient Descent), and structured academic resource roadmaps for developers and students.
Pure Python implementation of core linear algebra concepts from scratch, including vectors, matrices, determinants, inverses, and linear system solvers without using NumPy.
Pure Python & NumPy implementations of core supervised and unsupervised machine learning algorithms built from first principles.
Implementations of canonical machine learning algorithms from first principles using NumPy and Pandas.
Machine learning algorithms built from scratch in pure NumPy alongside optimized production pipelines
To associate your repository with the algorithms-from-scratch topic, visit your repo's landing page and select "manage topics."