On the Relationship Between Interpretability and Explainability in Machine Learning

TL;DR

This paper challenges the common view that interpretability and explainability are separate, arguing they are complementary and should be integrated to improve understanding of machine learning models.

Contribution

It critically examines the relationship between interpretability and explainability, advocating for combined approaches that leverage the strengths of both.

Findings

Interpretability and explainability are often viewed as substitutes, but are actually complementary.

Both approaches have unique shortcomings that can be mitigated by integrating them.

A call for research that targets both interpretability and explainability simultaneously.

Abstract

Interpretability and explainability have gained more and more attention in the field of machine learning as they are crucial when it comes to high-stakes decisions and troubleshooting. Since both provide information about predictors and their decision process, they are often seen as two independent means for one single end. This view has led to a dichotomous literature: explainability techniques designed for complex black-box models, or interpretable approaches ignoring the many explainability tools. In this position paper, we challenge the common idea that interpretability and explainability are substitutes for one another by listing their principal shortcomings and discussing how both of them mitigate the drawbacks of the other. In doing so, we call for a new perspective on interpretability and explainability, and works targeting both topics simultaneously, leveraging each of their respective assets.