// the find
fengdu78/WZU-machine-learning-course
温州大学《机器学习》课程资料（代码、课件等）
Course materials from Wenzhou University's machine learning curriculum, taught by Professor Huang Haiguang. Covers classical ML algorithms (regression through SVMs, clustering, dimensionality reduction) plus a deep learning track, all in Jupyter notebooks with matching PDF slides. Aimed at Chinese undergraduates new to the field, though the notebooks are readable enough to work through without the lectures.
The coverage is genuinely complete for a classical ML survey — linear regression to ensemble methods to association rules, each with its own self-contained notebook and real datasets. The deep learning supplement (15 PDFs) extends through Transformers and ViT, which is more current than most university course repos. Scipy coverage is unusually thorough: 10 separate notebooks for optimize, linalg, stats, signal, sparse, etc., which is rare to see treated as first-class prerequisite material. The numpy-100 exercises and pandas exercise sets are included in-tree, so you get drill problems alongside the lecture code.
Everything is in Chinese — slides, notebook prose, directory names — which hard-blocks non-Chinese readers even though the code itself is international. The notebooks are organized as lecture companions, not standalone tutorials; without the video or slides, several jump straight into code with minimal explanation of why choices were made. There's no requirements.txt or environment file, so you're on your own figuring out which scikit-learn and pandas versions the notebooks were written against. The deep learning section is all PDFs with no accompanying code, which is a significant gap given how much of DL understanding comes from running things.