1 paper · 2 filters
David A. Kelly, Nathan Blake
Explainability techniques are used to assess the output of various deep learning models. This is especially true in healthcare, where models need to be trusted and decisions justif…