ml-rigor - Apply rigorous evaluation to machine learning pipelines
Enforces baseline comparisons, cross-validation, model interpretation, calibration, and leakage prevention for machine learning pipelines.
Tags
Updated: 2026-10-16Capabilities
Typical Inputs
Typical Outputs
What this skill does
- Compare models against baselines
- Run stratified cross-validation
- Report score variance
- Calculate confidence intervals
- Tune hyperparameters with randomized search
- Estimate performance with nested CV
- Check probability calibration
- Apply probability calibration
- Prevent information leakage
- Flag exploratory findings
Inputs
- Training data
- Target labels
- Machine learning model
- Task type
- Evaluation metrics
- Domain heuristic
Outputs
- Baseline metrics
- Cross-validation metrics
- Confidence intervals
- Hyperparameter score distributions
- Calibration metrics
- Calibration curve figure
- Overfitting warnings
- Exploratory quality status
Requirements
- Python environment
- scikit-learn
- NumPy
- SciPy
- Matplotlib for calibration plots