Recent advances in explainable deep learning and how to model uncertainty in explainability.

Presenter: Kristoffer Knutsen Wickstrøm, PhD candidate at UiT –The Arctic University of Norway

Abstract

Deep learning is the cornerstone of artificial intelligence applications across a wide range of tasks and domains. An important component that is missing from deep learning is explainability, i.e. the ability to explain what influenced a prediction made by a deep learning-based system. Explainable deep learning is an active area of research, with new algorithms being proposed at a rapid pace. This presentation will give a brief review of existing methods for explainable deep learning, and a more in-depth presentation of some recent methods. Furthermore, modeling uncertainty in explainability has gained less attention but is also a crucial component to design trustworthy deep learning-based systems. Recent works on modeling uncertainty in explainability will be presented, which will be exemplified with applications to medical data.

‍