Explanation Relevance Index - A New Way of Assessing the Quality of Explanations Based on Learned Features.
Journal:
Studies in health technology and informatics
Published Date:
Oct 20, 2023
Abstract
The XAI methods began to emerge as a response for the black-box methods used to make decisions that could not be explained, even if checked by humans they were correct. This created the need for a better evaluation for these methods. In this paper we propose a new method for evaluation based on features. The main advantage of applying ERI to CNNs explanations are: a fully automated way to measure the quality of an explanation and the fact that the score uses the same information as the CNN, in this way being able to offer a measure of the quality of explanation that can be validated automatically.
