SHAP and Shapley Values Loop Nodes with a Custom Regression Model

This worflow shows how to use SHAP and Shapley Values Loop nodes and it creates Stacked Bar Charts similar to Force Plots for you to compare and understand the explanations. SHAP algorithm needs a smaller table to represent the validation set. This is achieved with the SHAP Summarizer Component.

This workflow showcases an example of using Gradient Boosted Trees Regression model, the workflow can be used with any other model as well.

External resources

• SHAP (SHapley Additive exPlanations) - Python Library on GitHub
• Explaining prediction models and individual predictions with feature contributions, Štrumbelj and Kononenko, 2014
• Verified Components
• SHAP Dependence Plot (documentation)