Expand description
Adaptive Threshold Learning System (PMAT-037)
Dynamic threshold learning that adjusts warning/critical bounds based on historical baseline data.
§Features
- Baseline learning from historical samples (μ±2σ)
- Percentile-based threshold computation
- Outlier filtering to prevent over-learning
- User override support for static thresholds
§Falsification Criteria (F1291-F1300)
See tests/adaptive_threshold_f1291.rs for falsification tests.
Structs§
- Learned
Threshold - Learned threshold bounds
- Threshold
Check - Result of threshold check
- Threshold
Learner - Threshold learner
Enums§
- Threshold
Direction - Threshold direction
Constants§
- DEFAULT_
CONFIDENCE_ LEVEL - Default confidence level for bounds (95%)
- DEFAULT_
OUTLIER_ THRESHOLD - Default outlier threshold (3 standard deviations)
- MIN_
SAMPLES_ FOR_ LEARNING - Minimum samples required for learning