Placeholder chapter page. Sections are produced by the book-skills pipeline.
Sections
- 17.1 Why labels are scarce in sensing
- 17.2 Contrastive learning with sensor augmentations
- 17.3 Masked reconstruction (MAE for time series)
- 17.4 Temporal predictive coding and TS2Vec
- 17.5 Relative and relational objectives (RelCon)
- 17.6 Cross-device, cross-user, cross-modal pretraining
- 17.7 Probing and evaluating learned representations
Lab 17
pretrain a contrastive/masked encoder on unlabeled sensor windows; fine-tune with few labels and measure the label-efficiency curve.