T6B — Time-series / HAR — sensor drift¶
Paper evidence: main.pdf · Block findings
Lemma: D6 · Stack: pytorch
Nuisance key: temporal
Production change: Sensor aging, device, session drift; activity label fixed.
Notebook (Run All, built-in demo): t06b-temporal-har.ipynb
pip install matching-pmh torch
# Open the notebook and Run All
What this task achieved (headline)¶
Matched PMH wins HAR stress 3.0: bal. acc 0.4099 vs baseline 0.2794 (3 seeds).
| baseline | PMH | wrong_W |
|---|---|---|
| 0.2794 @ stress 3 | 0.4099 | fails geometry |
Paper preset: t6_temporal_d6 · from pmh.benchmark.presets import get_preset
Subtasks (paper)¶
HAR multi-seed paper runs¶
PMH 0.4099 vs 0.2794 @ stress 3.
Preset: t6_temporal_d6
Collect W from baseline¶
Preset: t6_temporal_d6
Stress robustness eval¶
Preset: t6_temporal_d6
Run with matching-pmh¶
from pmh import PMHTrainer, evaluate_robust_fit
# nuisance="temporal"
Do not use PMH when¶
New activities only at deploy.
Replace demo data with yours¶
Swap demo loaders for your train_loader, source_batches, target_batches, and deploy holdout. Hook the backbone before your task head.