ljpzzz/machinelearning — reverse-engineered prompt

Build me a clean learning repository to go with a Chinese machine learning blog. I want it to feel like a study companion where readers can open a notebook or script, run the example, and connect it back to the article they just read.

Organize it by topic, including basic regression, classification, clustering, dimensionality reduction, ensemble learning, math and statistics, association rules, recommendation, natural language processing, reinforcement learning, and practical model deployment notes. Add runnable examples using common Python data science tools, with notebooks for classic machine learning demos and scripts for reinforcement learning examples like SARSA, Q Learning, DQN, policy gradient, actor critic, A3C, and DDPG.

Please write a clear Chinese README with article links on one side and code links on the other, plus notes about older Python 2.7 snippets and newer Python 3 examples. Keep the code simple and educational, not production heavy. Look up current docs online if you need to.