Fast Calculation of Feature Contributions in Boosting Trees

TL;DR

This paper introduces Q-SHAP, an efficient algorithm that significantly reduces the computational complexity of calculating feature contributions in boosting trees, enabling both faster and more accurate global feature importance evaluation.

Contribution

The paper presents Q-SHAP, a novel polynomial-time algorithm for computing Shapley values under quadratic loss, improving efficiency and accuracy over existing methods.

Findings

Q-SHAP reduces computational complexity to polynomial time.

Q-SHAP improves the accuracy of feature-specific R^2 estimates.

Simulations demonstrate enhanced efficiency and accuracy of Q-SHAP.

Abstract

Recently, several fast algorithms have been proposed to decompose predicted value into Shapley values, enabling individualized feature contribution analysis in tree models. While such local decomposition offers valuable insights, it underscores the need for a global evaluation of feature contributions. Although coefficients of determination ($R^2$) allow for comparative assessment of individual features, individualizing $R^2$ is challenged by the underlying quadratic losses. To address this, we propose Q-SHAP, an efficient algorithm that reduces the computational complexity of calculating Shapley values for quadratic losses to polynomial time. Our simulations show that Q-SHAP not only improves computational efficiency but also enhances the accuracy of feature-specific $R^2$ estimates.