A SHAP summary plot is a powerful visualization tool used to interpret the output of machine learning models. It provides a global view of how features influence predictions across the dataset.

Purpose

• Helps you understand which features matter most.

• Shows the direction of influence (positive/negative).

• Reveals feature interactions and variability.

What Are SHAP Values?

SHAP values are based on Shapley values from cooperative game theory. They attribute each feature's contribution to a particular prediction fairly and consistently.

For each prediction:

• The model’s output is broken down into the sum of the SHAP values of each feature + a base value (bias).

• Each SHAP value tells you how much a feature pushed the prediction up or down compared to the average prediction.