On the Relationship Between Interpretability and Explainability in Machine Learning
This work addresses the problem of improving transparency in ML for high-stakes decision-making, but it is incremental as it builds on existing dichotomous literature.
The paper challenges the common view that interpretability and explainability are substitutes in machine learning by analyzing their shortcomings and proposing they mitigate each other's drawbacks, advocating for a new perspective that integrates both.
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.