Feature engineering
Where most of the accuracy actually comes from on tabular problems, and where most production failures are born.
| # | File | The question it answers |
|---|---|---|
| 01 | 01_encoding_categoricals.md | how to encode a 40,000-value column without one-hot |
| 02 | 02_scaling_and_transforms.md | which models need scaling, skew, outliers, missing values |
| 03 | 03_feature_selection.md | which features to keep — including operational cost, not just statistics |
| 04 | 04_data_leakage.md | the most expensive bug in ML |
| 05 | 05_imbalanced_data.md | 99:1 classes without reaching for SMOTE first |
| 06 | 06_feature_stores.md | training/serving skew and point-in-time correctness |
Read 04 first. Leakage doesn’t crash, doesn’t warn, and makes your metrics better — which is exactly why it reaches production. The reflex to build is: unexpectedly good results are leakage until proven otherwise.