SimplEx is a user-centred method that provides example-based explanations with reference to a freely selected set of examples, called the corpus . SimplEx uses the corpus to improve the user’s understanding of the latent space with post-hoc explanations answering two questions: (1) Which corpus examples explain the prediction issued for a giventest example? (2) What features of these corpus examples are relevant for themodel to relate them to the test example? SimplEx provides an answer byconstructing the test latent representation as a mixture of corpus latentrepresentations . We propose a novel approach, the Integrated Jacobian, that allows SimplEx to make explicit the contribution of each corpus feature inthe mixture. Through experiments on tasks ranging from mortality prediction to image classification, we demonstrate that these decompositions are robust and accurate. With illustrative use cases in medicine, we show that SimplEx demonstrates how the freedom in choosing thecorpus allows the user to have personalized explanations in terms of examplesthat are meaningful for them .

Author(s) : Jonathan Crabbé, Zhaozhi Qian, Fergus Imrie, Mihaela van der Schaar

Links : PDF - Abstract

Code :
Coursera

Keywords : corpus - simplex - examples - user - latent -

Leave a Reply

Your email address will not be published. Required fields are marked *