this repository was planned, to have theory of every topic, but I dont get enough time to document everything. now only necessary and popular algorithms will be covered.
This reinforcement learning series is inspired by the Google DeepMind XUCL Reinforcement Learning YouTube series.
Everything in this repository is structured according to the content taught by Research Scientist Hado van Hasselt, Diana Borsa, and Research Engineer Matteo Hessel.
This repository is built for:
- Heavy Theory (even at research level)
- Core deep dive into algorithms
- Algorithm implementations
- Experiments
- Reference code
This repository is NOT meant to be a theory book.
All theory, learning roadmap, deep explanation, intuition, and structured lectures will be published on Substack.
If you are following this repository, do not start learning only from code.
The actual learning path will live on Substack. (with code explanations on Substack) This GitHub repository exists only to support learning through implementations.
Once Substack posts are available:
- You should read theory on Substack first (with code explainations there: to make it easy theory with code-- explained in simple english)
- Then return to GitHub for code and experiments (full code setup of algorithms and model's etc)
Learning happens on Substack.
Full Scale Implementation happens on GitHub.
Each topic in this repo contains two folders:
theoryalgorithms
- The theory folder will contain links to Substack articles (when published).
- The algorithms folder contains Python implementations.
Substack is used because it allows:
- Visual explanations
- Code walkthroughs
- Mathematical clarity
- Better structured learning than GitHub Markdown
GitHub is used for:
- Code
- Experiments
- Reproducible implementations
This repository is currently in progress.
Content, structure, and implementations will be added continuously.
This project is designed as:
Learn from Substack (deep dive at research level: with code explained in simple english, with visuals).
Practice on GitHub.
If you are serious about mastering Reinforcement Learning and Deep Reinforcement Learning, follow both together.
This repository is designed for everyone, including:
- High school students starting Reinforcement Learning
- College students learning AI / ML
- Data Scientists
- Machine Learning Engineers
- Software Developers
- Research Engineers
- AI Researchers
- Self-taught learners
- Anyone transitioning into AI / Deep Learning
The content is structured so that:
- Beginners can build foundations step-by-step
- Intermediate learners can strengthen understanding
- Advanced learners can implement research-level algorithms
- Researchers can experiment with modern methods
No matter your background, if you are serious about learning Reinforcement Learning, this repository is built to support you.