COMS W4721, Columbia University
Why another set of ML notes? Machine learning is a constantly evolving field. When I first taught this course, I did not assign readings or a textbook, and I made the lectures self-contained.
Still, when I have the time, I find that writing and revising notes helps me organize my learning and teaching, so here is my attempt.
Context. The notes were drafted by AI from my Spring 2026 lecture notes and recordings, with my input along the way. This is an evolving project that I will keep revising: the structure and topics may change, and the notes may contain mistakes. Students who plan to take the course in Spring 2027 should treat what I cover in lecture after the spring semester begins as the official reference.
Acknowledgment. This course is based on Nakul Verma’s machine learning course at Columbia, which I took as a student. From it, I learned not only machine learning but also how to teach it.