Joao Marques-Silva: Formal Reasoning Methods in Explainable AI

The forecasted applications of machine learning (ML) in safety critical applications hinge on systems that are robust in their operation and that can be trusted. This talk overviews recent efforts on applying automated reasoning tools in explaining non-interpretable (black-box) ML models. Moreover, the talk details the computation of rigorous explanations of black-box models, and how these serve for assessing the quality of widely used heuristic explanation approaches. The talk also covers important properties of rigorous explanations, namely duality properties between different kinds of explanation. Finally, the talk briefly overviews ongoing work on mapping tractable explainability.