Readings

Unordered collection of various resources I found particularly interesting.

Designing great data products: A legendary piece describing the biggest flaws of modern recommender systems and shining lights on things to consider when it comes explore vs exploit trade-offs
How to initialize deep neural networks? Xavier and Kaiming initialization: Most practitioners were using Xavier initialization as a default because it was baked into caffee libraries. Few people understood why until this question was raised openly.
The 1cycle policy: A good example of empiricists showcasing valuable insights from practical projects.
The decade of Deep Learning: A fantastic summary of 10 years of deep learning.
Lessons learned building an ML trading system that turned $5k into $200k: A pragmatic look at machine learning in financial systems.
L2 Regularization and Batch Norm: An often overlooked property of L2/batch norm interactions with neural networks.
The LASSO Estimator: My co-worker Andrew Tulloch goes deep about his love for LASSO estimators.
Gradient Boosting Machine: A solid look at the mathematics behind gradient boosting machines.
Modeling conversion rates using Weibull and gamma distributions: In-depth article about conversion rates/churn predictions

Practical Deep Learning for Coders (v3): With this course you will build from the ground up the state of the art deep learning architectures for tabular data, computer vision and NLP!
Practical Deep Learning for Coders (v1): When I get asked "how do I start/freshen up on ML? I answer with this.

Deep Learning Book: A thorough book that will guide you through many subtleties of training neural networks. This book covers everything from training to inference.
The Mechanics of Machine Learning: Jeremy Howard is a titan when it comes to ML education. This is a book co-authored with Terence Parr which constitutes a comprehensive introduction to machine learning practices.

Doctor vs Plumber: Which person is wealthier at Age 42: Great personal finance introductory video, contains all the basics clearly laid down.
Fermat's two square theorem: Mathologer is in my opinion one of the best Math Youtubers out there. This is his visual proof to Fermat's two square theorem.