On Efficiently Explaining Graph-Based Classifiers
This work addresses the need for interpretability in machine learning by providing efficient explanation methods for graph-based classifiers, which is incremental as it extends existing techniques to broader classifier types.
The paper tackles the problem of efficiently computing explanations for graph-based classifiers, such as decision trees and binary decision diagrams, by proposing polynomial-time algorithms for generating one PI-explanation and one contrastive explanation, with experimental validation on public benchmarks.
Recent work has shown that not only decision trees (DTs) may not be interpretable but also proposed a polynomial-time algorithm for computing one PI-explanation of a DT. This paper shows that for a wide range of classifiers, globally referred to as decision graphs, and which include decision trees and binary decision diagrams, but also their multi-valued variants, there exist polynomial-time algorithms for computing one PI-explanation. In addition, the paper also proposes a polynomial-time algorithm for computing one contrastive explanation. These novel algorithms build on explanation graphs (XpG's). XpG's denote a graph representation that enables both theoretical and practically efficient computation of explanations for decision graphs. Furthermore, the paper proposes a practically efficient solution for the enumeration of explanations, and studies the complexity of deciding whether a given feature is included in some explanation. For the concrete case of decision trees, the paper shows that the set of all contrastive explanations can be enumerated in polynomial time. Finally, the experimental results validate the practical applicability of the algorithms proposed in the paper on a wide range of publicly available benchmarks.