Abductive explanations of classifiers under constraints: Complexity and properties

TL;DR

This paper introduces new abductive explanation types for classifiers that incorporate feature constraints, reducing redundancy and analyzing their complexity and formal properties.

Contribution

It proposes three novel explanation types considering feature constraints, along with complexity analysis and formal guarantees, enhancing interpretability methods.

Findings

Coverage effectively discards redundant explanations

Complexity varies across explanation types

Catalogue of explanations with formal properties

Abstract

Abductive explanations (AXp's) are widely used for understanding decisions of classifiers. Existing definitions are suitable when features are independent. However, we show that ignoring constraints when they exist between features may lead to an explosion in the number of redundant or superfluous AXp's. We propose three new types of explanations that take into account constraints and that can be generated from the whole feature space or from a sample (such as a dataset). They are based on a key notion of coverage of an explanation, the set of instances it explains. We show that coverage is powerful enough to discard redundant and superfluous AXp's. For each type, we analyse the complexity of finding an explanation and investigate its formal properties. The final result is a catalogue of different forms of AXp's with different complexities and different formal guarantees.