Skip to content

feat: implement human based interfaces - #3

Merged
alexnodeland merged 2 commits into
mainfrom
feat/human-based
Dec 13, 2025
Merged

feat: implement human based interfaces#3
alexnodeland merged 2 commits into
mainfrom
feat/human-based

Conversation

@alexnodeland

Copy link
Copy Markdown
Owner

No description provided.

alexnodeland and others added 2 commits December 12, 2025 18:12
- Add proper Bradley-Terry MLE with Newton-Raphson and MM algorithms
  - Newton-Raphson provides analytical covariance from Fisher Information
  - MM algorithm uses bootstrap resampling for variance estimation
  - Both support convergence checking and regularization

- Add uncertainty quantification throughout the system
  - FitnessEstimate struct with mean, variance, and 95% CIs
  - WelfordVariance for numerically stable online variance calculation
  - Candidates now track fitness_with_uncertainty

- Add active learning selection strategies
  - Sequential: original round-robin behavior
  - UncertaintySampling: prioritizes high-variance candidates
  - ExpectedInformationGain: maximizes entropy reduction
  - CoverageAware: ensures minimum coverage before exploring

- Integrate uncertainty into session and algorithm
  - sync_fitness_estimates() to update candidates from aggregator
  - candidates_by_uncertainty() for identifying under-evaluated candidates
  - Selection strategy configurable via builder pattern

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
@alexnodeland
alexnodeland merged commit 17ee2e0 into main Dec 13, 2025
6 checks passed
@alexnodeland
alexnodeland deleted the feat/human-based branch December 13, 2025 01:08
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant