Consistent feature attribution for tree ensembles
This addresses a critical problem for users of gradient boosting trees and random forests by providing consistent feature importance explanations, though it is incremental as it builds on existing SHAP theory.
The paper tackled the inconsistency of current feature attribution methods for tree ensembles, where increasing a feature's reliance can decrease its assigned importance, and introduced fast exact solutions for SHAP values, resulting in significantly improved supervised clustering performance.
Note that a newer expanded version of this paper is now available at: arXiv:1802.03888 It is critical in many applications to understand what features are important for a model, and why individual predictions were made. For tree ensemble methods these questions are usually answered by attributing importance values to input features, either globally or for a single prediction. Here we show that current feature attribution methods are inconsistent, which means changing the model to rely more on a given feature can actually decrease the importance assigned to that feature. To address this problem we develop fast exact solutions for SHAP (SHapley Additive exPlanation) values, which were recently shown to be the unique additive feature attribution method based on conditional expectations that is both consistent and locally accurate. We integrate these improvements into the latest version of XGBoost, demonstrate the inconsistencies of current methods, and show how using SHAP values results in significantly improved supervised clustering performance. Feature importance values are a key part of understanding widely used models such as gradient boosting trees and random forests, so improvements to them have broad practical implications.