Placeholder chapter page. Sections are produced by the book-skills pipeline.
Sections
- 8.1 Time-domain features
- 8.2 Frequency- and time-frequency features
- 8.3 Statistical, shape, and entropy features
- 8.4 Domain-specific feature libraries (tsfresh, catch22)
- 8.5 Feature selection
- 8.6 PCA, ICA, and manifold methods (UMAP, t-SNE) for sensors
- 8.7 When handcrafted features still beat deep learning
Lab 8
build a feature-based activity/condition classifier; compare against a small neural baseline and analyze where each wins.