04 — Machine Learning¶
Classical machine learning: scikit-learn, preprocessing, model selection, evaluation, feature engineering, clustering, dimensionality reduction.
Scope¶
Cheat sheets for traditional ML algorithms, pipelines, and evaluation methodology. Deep learning belongs to 05-Deep-Learning.
Cheat Sheets¶
| Sheet | Covers |
|---|---|
| scikit-learn | Estimator API, pipelines, train/test, tuning, metrics, persistence |
| ml-preprocessing | Scaling, encoding, imputation, ColumnTransformer, leakage |
| ml-model-selection | Splits, k-fold CV, Grid/RandomizedSearchCV, nested CV |
| ml-evaluation | Confusion matrix, precision/recall/F1, ROC/PR, regression metrics |
| ml-feature-engineering | Derived features, dates/text, target encoding, selection |
| ml-clustering | KMeans, DBSCAN, hierarchical, choosing k, segmentation |
| ml-dimensionality-reduction | PCA, t-SNE, UMAP, LDA, explained variance |
Contribute¶
- Template: CHEATSHEET_TEMPLATE.md
- Guide: CONTRIBUTING.md
- Planned: ensemble methods deep-dive, XGBoost/LightGBM, time-series models