Is Machine-Learning-Collection open source?

Yes — aladdinpersson/Machine-Learning-Collection is open source, released under the MIT license.

What language is Machine-Learning-Collection written in?

aladdinpersson/Machine-Learning-Collection is primarily written in Python.

How popular is Machine-Learning-Collection?

aladdinpersson/Machine-Learning-Collection has 8.5k stars on GitHub.

Where can I find Machine-Learning-Collection?

aladdinpersson/Machine-Learning-Collection is on GitHub at https://github.com/aladdinpersson/Machine-Learning-Collection.

← all repositories

aladdinpersson/Machine-Learning-Collection

The anti-course: a lookup-style repo for ML algorithms and PyTorch

Aladdin Persson open-sourced his video tutorial code so developers can copy, tweak, and lookup ML implementations from linear regression to Transformers.

★8.5k stars Python Learning ML Frameworks

View on GitHub ↗ Homepage ↗

Not currently ranked — collecting fresh signals.

star history

What it does

This repository is a curated grab bag of machine learning tutorials and reference implementations, spanning classic algorithms written from scratch and modern deep learning recipes in PyTorch and TensorFlow. The author organizes it as a lookup resource rather than a linear course, with each topic linking to a dedicated YouTube walkthrough. You’ll find everything from gradient-descent linear regression and CVXOPT-based SVMs to YOLOv3, DCGANs, and Transformers built from scratch.

The interesting bit

The repo treats YouTube as a first-class citizen: nearly every script has a companion video, making it a rare hybrid of visual lecture and copy-pasteable code. The scope is also unusually wide, jumping from scratch-built decision trees and k-means to full-blown object detection and GANs without pretending to be a single coherent framework.

Key highlights

Classic ML from scratch: linear regression (gradient descent and normal equation), logistic regression, Gaussian naive Bayes, k-NN, k-means, SVM via CVXOPT, neural networks, and decision trees.
Extensive PyTorch coverage: tensor basics, CNNs, RNNs/GRUs/LSTMs, bidirectional RNNs, custom datasets, mixed precision, transfer learning, data augmentation, TensorBoard, and weight initialization.
Advanced deep learning: U-Net segmentation, neural style transfer, image captioning, Seq2Seq with attention, Transformers from scratch, and YOLOv1/v3 object detection.
GAN progression: simple fully-connected GAN, DCGAN, and WGAN implementations.
TensorFlow tutorials for beginners and CNN architectures, though the catalog is noticeably smaller than the PyTorch side.

Caveats

The repo is heavily tied to YouTube; the code is written to accompany video walkthroughs, so context may be thin without watching.
TensorFlow content is limited to beginner tutorials and CNN architectures, while PyTorch dominates the advanced sections.
The README is a flat table of contents, so learners seeking a structured curriculum will need to chart their own path.

Verdict

Ideal for developers who learn by reading code and want a broad, searchable index of ML techniques from scratch implementations to production-flavored deep learning. Less useful if you need a structured course or a monolithic library with unified APIs.

Frequently asked

What is aladdinpersson/Machine-Learning-Collection?: Aladdin Persson open-sourced his video tutorial code so developers can copy, tweak, and lookup ML implementations from linear regression to Transformers.
Is Machine-Learning-Collection open source?: Yes — aladdinpersson/Machine-Learning-Collection is open source, released under the MIT license.
What language is Machine-Learning-Collection written in?: aladdinpersson/Machine-Learning-Collection is primarily written in Python.
How popular is Machine-Learning-Collection?: aladdinpersson/Machine-Learning-Collection has 8.5k stars on GitHub.
Where can I find Machine-Learning-Collection?: aladdinpersson/Machine-Learning-Collection is on GitHub at https://github.com/aladdinpersson/Machine-Learning-Collection.