An Empirical Study of Explainable AI Techniques on Deep Learning Models For Time Series Tasks
Machine Learning
2020-12-09 v1 Artificial Intelligence
Abstract
Decision explanations of machine learning black-box models are often generated by applying Explainable AI (XAI) techniques. However, many proposed XAI methods produce unverified outputs. Evaluation and verification are usually achieved with a visual interpretation by humans on individual images or text. In this preregistration, we propose an empirical study and benchmark framework to apply attribution methods for neural networks developed for images and text data on time series. We present a methodology to automatically evaluate and rank attribution techniques on time series using perturbation methods to identify reliable approaches.
Cite
@article{arxiv.2012.04344,
title = {An Empirical Study of Explainable AI Techniques on Deep Learning Models For Time Series Tasks},
author = {Udo Schlegel and Daniela Oelke and Daniel A. Keim and Mennatallah El-Assady},
journal= {arXiv preprint arXiv:2012.04344},
year = {2020}
}
Comments
5 Pages, 2 Pages References, Pre-registration workshop NeurIPS (2020), Vancouver, Canada