Courses, textbooks, and foundations
Find another explanation or study a foundation systematically. Use Crash Course for the big picture, 3Blue1Brown for mathematical intuition, and Mu Li to connect models to code.
Computing: what computers and programs do
Crash Course Computer Science — Short English videos hosted by Carrie Anne Philbin. Start with binary, CPUs, programs, operating systems, files, and the internet to build an overall picture. Official introduction
CS Self-Learning Guide — Chinese course navigation and learning experience. Read the introduction and usage guide, then find a course for the subject you need.
The Missing Semester of Your CS Education — English lectures and exercises on everyday tools. Start with the shell, Git, and debugging to learn command-line work, version control, and diagnosis.
Pro Git, Chinese edition — An online book. Begin with basic operations, history, and remote collaboration; study branches when working with others.
Mathematics: intuition, foundations, and depth
3Blue1Brown: Linear Algebra — Visual English explanations of vectors, transformations, matrix multiplication, and eigenvectors. The geometry helps explain why the formulas look as they do. Bilibili channel
3Blue1Brown: The Essence of Calculus — English videos and illustrated text. Build intuition for rates of change, small changes, and integration, then connect gradients and the chain rule.
Seeing Theory — Brown University — Interactive probability and statistics. Change parameters and inspect distributions and samples to understand the concepts.
MIT 18.06 Linear Algebra — Gilbert Strang — English lectures, notes, and assignments from 2010. Study subspaces, orthogonality, eigenvalues, and SVD when you need systematic foundations.
Harvard Stat 110 — Joe Blitzstein — English lectures, textbook, and exercises. Conditional probability, expectation, and common distributions recur in ML and RL; pair concepts with exercises.
Mathematics for Machine Learning — An open English textbook connecting mathematics to ML. Consult matrix decompositions, vector calculus, probability, and optimization.
Machine learning and deep learning
An Introduction to Statistical Learning — English textbook and courses with Python and R editions. Start with training, generalization, regression, and classification, then model selection and resampling; choose algorithms by task later.
Dive into Deep Learning — Chinese textbook and code. Work through data operations, linear models, losses, optimization, and multilayer perceptrons, then choose architectures by interest. Official course index
Hung-yi Lee's 2025 Machine Learning Course — Chinese explanations, slides, and videos using generative AI to explain ML concepts. Select the lectures relevant to your question.
Stanford CS229 — A systematic English course on algorithm assumptions, derivations, and statistical learning. Notes can also be consulted by question.
Practical Deep Learning for Coders — fast.ai — An English practical course for people with some programming background. Build a working model first, then understand the components; begin with lesson one and its exercises.
3Blue1Brown: But what is a Neural Network? — Visual videos and text. Build a picture of inputs, layers, parameters, and outputs, then connect gradient descent and backpropagation.
StatQuest with Josh Starmer — Short English videos offering another explanation of cross-validation, bias and variance, tree models, and other statistics or ML concepts.
Python, tensors, and your first training loop
CS50's Introduction to Programming with Python — An English beginner course with exercises. Move from functions, conditions, and loops to exceptions, libraries, and files; practice explaining and changing small programs.
NumPy: the absolute basics for beginners — Official English introduction to array computing. Focus on shapes, indexing, axes, and broadcasting.
PyTorch: Learn the Basics — Official English tutorial connecting data, models, autograd, optimization, and saving into a training process. Useful for reading your first training loop.
Neural Networks: Zero to Hero — Karpathy — English videos and code requiring Python and basic mathematics. Begin with micrograd to see how gradients are computed, then move into language models.
Architectures: diagrams, code, and derivations
The Illustrated Transformer — Jay Alammar — English diagrams connecting token representations, attention, encoders, and decoders. Trace inputs and outputs before returning to equations.
The Annotated Transformer — Harvard NLP — Annotated English code and notebooks pairing the original Transformer paper with implementation. Useful after diagrams when you want to understand tensors and modules.
Transformer Explainer — Polo Club — Interactive visualization of inputs moving through a Transformer and how components affect next-token prediction.
Understanding LSTM Networks — colah — An illustrated English article from 2015. Learn LSTMs through states, gates, and sequence processing; useful alongside time-series models.
Scientific Spaces: article archive — Jianlin Su — Chinese technical archive. Follow the Transformer and diffusion series for positional encodings, architectures, and mathematical derivations.
Crash Course AI: an introduction to the field
Official course index · Episode 1: What Is Artificial Intelligence? · Preview
A short English video series introducing ML, language, RL, and robotics. Watch the opening episodes, then choose topics by interest. Continue to a small project to meet training, data, and evaluation in practice.