// the find
fengdu78/Coursera-ML-AndrewNg-Notes
吴恩达老师的机器学习课程个人笔记
Chinese-language study notes for Andrew Ng's 2014 Coursera ML course — all 10 weeks, translated and annotated, with Python rewrites of the original Octave programming exercises. Aimed squarely at Chinese-speaking developers who want a readable reference alongside the video lectures. Not a library, not a tool; purely a learning resource.
Complete coverage across all 18 lectures with a clean table of contents that lets you jump directly to any topic. Python Jupyter notebooks for all 8 exercise sets are a genuine improvement over the original Octave assignments — more people can run them without setup pain. Multiple export formats (markdown, HTML, docx, PDF) means it works offline, in editors, and in browsers. The 37k stars reflect real usage: this has been a go-to reference for Chinese ML students for years.
It documents the 2014 version of the course. Andrew Ng replaced that curriculum entirely in 2022 with a three-course specialization covering scikit-learn, decision trees with XGBoost, and modern neural network tooling — none of which is here. Math formulas in the HTML version are rendered as static images rather than MathJax or KaTeX, so they're unsearchable and break if the image host changes. The recent decision_tree addition appears to be from the new specialization (C2_W4 filename) grafted on without integrating it into the existing structure, which is a sign the repo is drifting rather than being maintained coherently. Anyone starting fresh should probably just use the updated course materials directly.