Active learning
Active learning skill. Uncertainty sampling, Query-by-Committee, expected model change, pool-based/stream-based, batch active learning, GP-based active learning with ARD-RBF kernel, dimension-adaptive convergence, stopping criteria, and model improvement pipeline.
Use This Skill When
- Uncertainty sampling.
- Query-by-Committee.
- Expected model change.
- Pool-based/stream-based.
- Batch active learning.
Required Inputs
- Research objective, decision target, or hypothesis.
- Available data, source constraints, and domain assumptions.
- Required outputs, success metrics, and deadline or reproducibility constraints.
Workflow
- Confirm scope, assumptions, and the exact artifact set to save.
- Apply the narrowest domain method that answers the request with defensible evidence.
- Save code, tables, figures, and intermediate outputs to files instead of chat-only output.
- State limitations, uncertainty, and any validation or sensitivity checks performed.
- Append skill selection, handoff I/O, and file writes to
logs/process-log.jsonl.
Deliverables
report.md: concise method, results, interpretation, and file inventory in the user's language.results/: structured outputs, metrics, model artifacts, or extracted findings.figures/: English-only charts, diagrams, or panels when visual output is needed.data/: processed or derived datasets when transformation occurs.
Quality Gates
- The selected method matches the scientific question and stated assumptions.
- Outputs are reproducible, saved to files, and traceable from inputs to conclusions.
- Missing data, uncertainty, bias, and hard limits are made explicit.
report.md and logs/process-log.jsonl reference the generated artifacts.
1---2name: scientific-active-learning3description: Active learning skill. Uncertainty sampling, Query-by-Committee, expected model change, pool-based/stream-based, batch active learning, GP-based active learning with ARD-RBF kernel, dimension-adaptive convergence, stopping criteria, and model improvement pipeline.4---56# Active learning78Active learning skill. Uncertainty sampling, Query-by-Committee, expected model change, pool-based/stream-based, batch active learning, GP-based active learning with ARD-RBF kernel, dimension-adaptive convergence, stopping criteria, and model improvement pipeline.910## Use This Skill When1112- Uncertainty sampling.13- Query-by-Committee.14- Expected model change.15- Pool-based/stream-based.16- Batch active learning.1718## Required Inputs1920- Research objective, decision target, or hypothesis.21- Available data, source constraints, and domain assumptions.22- Required outputs, success metrics, and deadline or reproducibility constraints.2324## Workflow25261. Confirm scope, assumptions, and the exact artifact set to save.272. Apply the narrowest domain method that answers the request with defensible evidence.283. Save code, tables, figures, and intermediate outputs to files instead of chat-only output.294. State limitations, uncertainty, and any validation or sensitivity checks performed.305. Append skill selection, handoff I/O, and file writes to `logs/process-log.jsonl`.3132## Deliverables3334- `report.md`: concise method, results, interpretation, and file inventory in the user's language.35- `results/`: structured outputs, metrics, model artifacts, or extracted findings.36- `figures/`: English-only charts, diagrams, or panels when visual output is needed.37- `data/`: processed or derived datasets when transformation occurs.3839## Quality Gates4041- The selected method matches the scientific question and stated assumptions.42- Outputs are reproducible, saved to files, and traceable from inputs to conclusions.43- Missing data, uncertainty, bias, and hard limits are made explicit.44- `report.md` and `logs/process-log.jsonl` reference the generated artifacts.
nahisaho