shap - Explain machine learning predictions with SHAP
Explain model predictions, compute feature attributions, generate SHAP visualizations, and analyze model behavior, bias, fairness, and feature interactions.
Tags
Updated: 2026-10-05Capabilities
Typical Inputs
Typical Outputs
What this skill does
- Compute SHAP values
- Explain individual predictions
- Visualize feature importance
- Debug model behavior
- Analyze bias and fairness
- Compare model explanations
- Analyze feature interactions
- Deploy explanation services
Inputs
- Trained machine learning model
- Feature data
- Background data
- Target labels
- Protected group attributes
Outputs
- SHAP values
- Feature importance plots
- Prediction explanations
- Model behavior analyses
- Fairness and bias analyses
- Explanation service endpoints
Requirements
- Python environment
- SHAP library
- Supported model framework