Saved in:
Bibliographic Details
Main Authors: Marques-Silva, Joao, Huang, Xuanxiang
Format: Preprint
Published: 2023
Subjects:
Online Access:https://arxiv.org/abs/2307.07514
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909099860426752
author Marques-Silva, Joao
Huang, Xuanxiang
author_facet Marques-Silva, Joao
Huang, Xuanxiang
contents Explainable artificial intelligence (XAI) aims to help human decision-makers in understanding complex machine learning (ML) models. One of the hallmarks of XAI are measures of relative feature importance, which are theoretically justified through the use of Shapley values. This paper builds on recent work and offers a simple argument for why Shapley values can provide misleading measures of relative feature importance, by assigning more importance to features that are irrelevant for a prediction, and assigning less importance to features that are relevant for a prediction. The significance of these results is that they effectively challenge the many proposed uses of measures of relative feature importance in a fast-growing range of high-stakes application domains.
format Preprint
id arxiv_https___arxiv_org_abs_2307_07514
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Explainability is NOT a Game
Marques-Silva, Joao
Huang, Xuanxiang
Artificial Intelligence
Machine Learning
Explainable artificial intelligence (XAI) aims to help human decision-makers in understanding complex machine learning (ML) models. One of the hallmarks of XAI are measures of relative feature importance, which are theoretically justified through the use of Shapley values. This paper builds on recent work and offers a simple argument for why Shapley values can provide misleading measures of relative feature importance, by assigning more importance to features that are irrelevant for a prediction, and assigning less importance to features that are relevant for a prediction. The significance of these results is that they effectively challenge the many proposed uses of measures of relative feature importance in a fast-growing range of high-stakes application domains.
title Explainability is NOT a Game
topic Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2307.07514