CS229 - Machine Learning

Machine Learning is the Stanford Engineering Everywhere release of Andrew Ng's graduate course CS229, a broad introduction to machine learning and statistical pattern recognition. The syllabus covers supervised learning (generative and discriminative models, parametric and non-parametric methods, neural networks, support vector machines), unsupervised learning (clustering, dimensionality reduction, kernel methods), learning theory (bias and variance trade-offs, VC theory, large margins) and reinforcement learning and adaptive control, and the lectures discuss applications from robotic control and data mining to speech recognition and text and web processing. The page asks for three prerequisites: enough computer science to write a non-trivial program, basic probability theory and basic linear algebra.

SEE provides 20 lecture videos with transcripts, the lecture notes and handouts, problem sets and additional resources. Access is free without registration under a CC BY-NC-SA 4.0 licence; nothing is graded and no credit or certificate is offered. The recording predates the deep-learning era, so it is strongest on fundamentals.