High Dimensional Model Explanations: an Axiomatic Approach

TL;DR

This paper introduces an axiomatic high-dimensional explanation method for black-box models that captures joint feature effects and is justified as the optimal local polynomial approximation.

Contribution

It proposes a novel axiomatization and a generalized Banzhaf index-based explanation method for high-dimensional models, addressing limitations of existing feature importance measures.

Findings

The method captures joint feature effects effectively.

It satisfies natural axioms and is the optimal local approximation.

Empirical results show it behaves predictably compared to other measures.

Abstract

Complex black-box machine learning models are regularly used in critical decision-making domains. This has given rise to several calls for algorithmic explainability. Many explanation algorithms proposed in literature assign importance to each feature individually. However, such explanations fail to capture the joint effects of sets of features. Indeed, few works so far formally analyze high-dimensional model explanations. In this paper, we propose a novel high dimension model explanation method that captures the joint effect of feature subsets. We propose a new axiomatization for a generalization of the Banzhaf index; our method can also be thought of as an approximation of a black-box model by a higher-order polynomial. In other words, this work justifies the use of the generalized Banzhaf index as a model explanation by showing that it uniquely satisfies a set of natural desiderata and that it is the optimal local approximation of a black-box model. Our empirical evaluation of our measure highlights how it manages to capture desirable behavior, whereas other measures that do not satisfy our axioms behave in an unpredictable manner.