A DIY Master's Syllabus for Machine Learning, Heavy on the Math

What it does This repository is a Chinese-language study guide that lays out a semester-style progression from undergraduate calculus to graduate-level machine learning. It assembles open courseware from MIT, Stanford, NTU, and Caltech into a single syllabus, pairing each lecture series with textbooks, notes, and explicit homework requirements. Think of it as a degree checklist without the tuition.

The interesting bit The authors do not treat programming as the main event; Python and Matlab are relegated to a brief “just enough to finish the assignments” sidebar. The real focus is on mathematical maturity—convex optimization and matrix theory are mandatory waypoints, not optional electives.

Key highlights

• Structured as a staged pipeline: basic math → programming → introductory ML → intermediate math → intermediate ML.
• Explicitly demands completing all instructor-assigned exercises and reading recommended papers, not just watching lectures.
• Difficulty-tagged book list separating “general” texts like An Introduction to Statistical Learning from harder classics like Pattern Recognition and Machine Learning.
• Points to follow-up repositories for specializations including deep learning, graphical models, and reinforcement learning.
• Heavy emphasis on foundational math: single and multivariable calculus, linear algebra, probability, and convex optimization.

Caveats

• Several entries contain blank or “暂无” (not available) slots for references and materials.
• It is purely a curated index of external links and reading lists; there is no original code or coursework here.
• Some programming sections are sparse, reflecting the authors’ view that coding is secondary to algorithmic intuition.

Verdict Ideal for self-learners with patience and a high tolerance for chalkboard math who want a credible alternative to a formal MS curriculum. Skip it if you are looking for hands-on notebooks or quick API tutorials.