Builds a complete, documented machine learning pipeline following the CRISP-DM framework (as taught in IS 455 at BYU). Use this skill whenever the user wants to build an ML pipeline, train a predictive model, do supervised learning, or work through business understanding, data understanding, data cleaning, modeling, or evaluation steps — even if they don't say "CRISP-DM" explicitly. Handles both classification and regression. Outputs either a Jupyter notebook (.ipynb) with markdown-documented phases or a fully commented Python script (.py).
此技能包含 SKILL.md 智能体指令文件。