Finding the right XAI method -- A Guide for the Evaluation and Ranking of Explainable AI Methods in Climate Science
Authors:
- Marina M. -C. HöhneMHMarina M. -C. Höhne
Preprint
Abstract
Explainable artificial intelligence (XAI) methods shed light on the predictions of machine learning algorithms. Several different approaches exist and have already been applied in climate science. However, usually missing ground truth explanations complicate their evaluation and comparison, subsequently impeding the choice of the XAI method. Therefore, in this work, we introduce XAI evaluation in the climate context and discuss different desired explanation properties, namely robustness, faithfulness, randomization, complexity, and localization. To this end, we chose previous work as a case study where the decade of annual-mean temperature maps is predicted. After training both a multi-layer perceptron (MLP) and a convolutional neural network (CNN), multiple XAI methods are applied and their skill scores in reference to a random uniform explanation are calculated for each property. Independent of the network, we find that XAI methods Integrated Gradients, layer-wise relevance propagation, and input times gradients exhibit considerable robustness, faithfulness, and complexity while sacrificing randomization performance. Sensitivity methods -- gradient, SmoothGrad, NoiseGrad, and FusionGrad, match the robustness skill but sacrifice faithfulness and complexity for randomization skill. We find architecture-dependent performance differences regarding robustness, complexity and localization skills of different XAI methods, highlighting the necessity for research task-specific evaluation. Overall, our work offers an overview of different evaluation properties in the climate science context and shows how to compare and benchmark different explanation methods, assessing their suitability based on strengths and weaknesses, for the specific research problem at hand. By that, we aim to support climate researchers in the selection of a suitable XAI method.Keywords
All Related Versions
This publication has 49 references indexed in Scilit:
- Evaluating feature attribution methods in the image domainMachine Learning, 2024
- This looks More Like that: Enhancing Self-Explaining Models by Prototypical Relevance PropagationPattern Recognition, 2022
- Interpretable and explainable AI (XAI) model for spatial drought predictionScience of The Total Environment, 2021
- Explainable Artificial Intelligence (XAI): Concepts, taxonomies, opportunities and challenges toward responsible AIInformation Fusion, 2020
- A deep learning approach to detecting volcano deformation from satellite imagery using synthetic datasetsRemote Sensing of Environment, 2019
- Towards a more reliable historical reanalysis: Improvements for version 3 of the Twentieth Century Reanalysis systemQuarterly Journal of the Royal Meteorological Society, 2019
- Methods for interpreting and understanding deep neural networksDigital Signal Processing, 2018
- Top-Down Neural Attention by Excitation BackpropInternational Journal of Computer Vision, 2017
- Explaining nonlinear classification decisions with deep Taylor decompositionPattern Recognition, 2017
- A machine learning nowcasting method based on real‐time reanalysis dataJournal of Geophysical Research: Atmospheres, 2017