What it does This repo is the official companion code for a Chinese MOOC called Python3 入门机器学习 (“Play with Machine Learning Algorithms”) on the Imooc platform. It contains chapter-by-chapter Jupyter notebooks plus a small custom library, playML, that re-implements classic algorithms from scratch in plain Python/NumPy.

The interesting bit The author doesn’t just import sklearn and move on. Many chapters include a parallel “our own implementation” track—kNN, linear regression, gradient descent, PCA—so you can see the math unboxed before the library call. It’s courseware with the training wheels partially removed.

Key highlights

• Covers kNN, linear regression, gradient descent variants, PCA, and touches on MNIST and eigenfaces
• Each topic pairs theory notebooks with runnable playML package code
• Includes NumPy/Matplotlib fundamentals for students new to the Python data stack
• Some supplementary sections marked “整理中” (organizing) / “敬请期待” (stay tuned)
• Author’s personal website noted as “废弃重整中” (abandoned, reorganizing)

Caveats

• README is entirely in Chinese; code comments and notebooks likely follow suit
• Several promised supplements are unfinished placeholders
• No English translation or community localization visible

Verdict Worth a look if you read Chinese and want to see algorithms built from the ground up alongside scikit-learn usage. Skip it if you need polished, self-contained English documentation or advanced deep-learning content.