Implement odds-based sequential decisions #26
Merged
wasabi opened 2 weeks ago

Implements sequential evidence and optimal-stopping improvements inspired by the Odds Algorithm.\n\nPhases:\n- Phase 0: fixes anomaly value extraction, sensor false-stability, typed scale thresholds, and enforceable solution gate evaluations.\n- Phase 1: adds immutable decision hypotheses and provenance-aware evidence events with dependence groups and likelihood inputs.\n- Phase 2: adds posterior odds updates, dependence rejection, finite-sequence odds strategy, utility-derived action thresholds, and threshold evaluations.\n- Phase 3: adds sequential experiment analysis with effect estimates, uncertainty bounds, benefit probability, futility, and safety stopping.\n- Phase 4: connects posterior decisions to strategy review tasks and solution gate evaluation.\n\nVerification:\n- Focused suite: 77 passed.\n- Full CI: 109 passed, 0 failed.\n- Migrations through schema 296 applied successfully.

Commits were merged into target branch
1/1
Submitter wasabi
Target main
Source feat/odds-sequential-decisions
Jobs
Merge Strategy
Create Merge Commit
Watchers (1)
Reference
pull request KI-26
Please wait...
Connection lost or session expired, reload to recover
Page is in error, reload to recover