Explaining predictive models with mixed features using Shapley values and conditional inference trees

TL;DR

This paper introduces a novel approach to explain mixed feature types in predictive models by modeling their dependence with conditional inference trees, improving interpretability over existing methods.

Contribution

It extends Shapley value explanations to mixed, dependent features using conditional inference trees, outperforming current industry standards in simulations and real data.

Findings

Our method often outperforms existing approaches in simulations.

It provides more accurate explanations for mixed feature types.

Application to financial data demonstrates practical effectiveness.

Abstract

It is becoming increasingly important to explain complex, black-box machine learning models. Although there is an expanding literature on this topic, Shapley values stand out as a sound method to explain predictions from any type of machine learning model. The original development of Shapley values for prediction explanation relied on the assumption that the features being described were independent. This methodology was then extended to explain dependent features with an underlying continuous distribution. In this paper, we propose a method to explain mixed (i.e. continuous, discrete, ordinal, and categorical) dependent features by modeling the dependence structure of the features using conditional inference trees. We demonstrate our proposed method against the current industry standards in various simulation studies and find that our method often outperforms the other approaches. Finally, we apply our method to a real financial data set used in the 2018 FICO Explainable Machine Learning Challenge and show how our explanations compare to the FICO challenge Recognition Award winning team.