shap - Explain machine learning predictions with SHAP
Explain machine learning predictions, feature importance, model behavior, bias, fairness, and feature interactions with SHAP.
Tags
Updated: 2026-10-05Capabilities
Typical Inputs
Typical Outputs
What this skill does
- Compute SHAP values
- Generate SHAP plots
- Explain individual predictions
- Analyze model behavior
- Analyze bias and fairness
- Compare model explanations
- Implement explanation services
Inputs
- Trained machine learning model
- Evaluation data
- Background data
- Target labels
- Protected attributes
Outputs
- SHAP value explanations
- Feature importance visualizations
- Individual prediction explanations
- Model debugging results
- Bias and fairness analysis
- Model comparison results
- Explanation service endpoints
Requirements
- SHAP-compatible model
- SHAP library
- Supported model framework