Over the course of my Computer Science and Data Science studies, I've come across some really valuable resources I believe have high SNR. I thought I'd share my list with other people who are interested in becoming ML engineers or students who are also enrolled in some related program.

Also, this list will stay updated so I'll add new things I find helpful over time.

General/Basic Resources and Overwiews

  • MLU Explain: Nice high-level ML principles explained in a visual way.
  • Data Driven Science & Engineering: A book going over the most important principles of Algebra and Calculus for ML/DS.
  • An Introduction to Statistical Learning: The holy grail of Statistical Learning. This book is really good for beginners since it covers almost all the basics that more complex types of models later build upon.
    • Stanford STATS202: Accompanying to the book, this syllabus covers the main topics with quick notes on how each algorithm works. Used this heavily when revising for the Statistical Learning class final.
  • Foundations of Machine Learning: A lot of alpha in this one. Fairly similar to ISLP. If you understand the concepts in here you can follow and understand every ML paper. Not kidding.
  • Andrej Karpathy on YouTube: You probably already know this one but I can't stress it enough; Karpathy & Chill is what gets you straight A's.
  • MIT 18.06 Linear Algebra: This playlist is a gem for anyone who's new to university level algebra. The videos are a bit older but Gilbert Strang's way of explaining concepts is second to none.
  • Cornell CS4780 ML: This playlist also covers the same subjects as An Introduction to Statistical Learning. I found Kilian Weinberger's lectures really good; he often brings a different perspective to some of the algorithms.

More in-depth

  • Speech and Language Processing Book: If you're new to natural language processing this is a good starting point. It covers pretty much everything up until Transformers.
  • An even more annotated Transformer: Amazing post on how Transformer Architectures work.
  • LLM Visualization: An amazing 3D map of the GPT architectures, its subcomponents and how everything is connected together.
  • The Novice's LLM Training Guide: Contains everything you need to know to train an LLM.
  • Yannic Kilcher: Yannic Kilcher posts videos on some larger and hot SOTA papers around every 2 weeks and it's a joy watching those to stay up to date.
  • Deep Learning: An MIT Press book: This is also an MIT release which goes in depth on the math used in deep learning. I really like this one because it tries to be straightforward and on point, little to no noise.

Finally, just looking at these resources won't be enough. Build stuff, break things and learn.