Efficient Contrastive Explanations on Demand

TL;DR

This paper introduces efficient algorithms for computing contrastive explanations in complex ML models by leveraging adversarial robustness, significantly improving explanation computation performance.

Contribution

It presents novel algorithms that compute contrastive explanations efficiently for models with many features, utilizing adversarial robustness insights.

Findings

Significant performance improvements over existing methods

Effective listing of multiple explanations

Ability to find smallest contrastive explanations

Abstract

Recent work revealed a tight connection between adversarial robustness and restricted forms of symbolic explanations, namely distance-based (formal) explanations. This connection is significant because it represents a first step towards making the computation of symbolic explanations as efficient as deciding the existence of adversarial examples, especially for highly complex machine learning (ML) models. However, a major performance bottleneck remains, because of the very large number of features that ML models may possess, in particular for deep neural networks. This paper proposes novel algorithms to compute the so-called contrastive explanations for ML models with a large number of features, by leveraging on adversarial robustness. Furthermore, the paper also proposes novel algorithms for listing explanations and finding smallest contrastive explanations. The experimental results demonstrate the performance gains achieved by the novel algorithms proposed in this paper.