Placeholder chapter page. Sections are produced by the book-skills pipeline.
Sections
- 5.1 Formats for telemetry, waveforms, event streams, and point clouds
- 5.2 Window construction and labeling
- 5.3 Leakage in sensor datasets (the field's most common error)
- 5.4 Device, user, site, session, and machine splits
- 5.5 Normalization and per-device calibration
- 5.6 Annotation quality, weak labels, and label delay
- 5.7 Reproducible, versioned sensor pipelines and data contracts
Lab 5
build leakage-safe subject- and device-disjoint splits for human activity recognition; measure the accuracy gap vs a naive random split.