A Consistent and Efficient Evaluation Strategy for Attribution Methods

With a variety of local feature attribution methods being proposed in recent years, follow-up work suggested several evaluation strategies. To assess the attribution quality across different attribution techniques, the most popular among these evaluation strategies in the image domain use pixel pert...

Full description

Saved in:
Bibliographic Details
Published in:arXiv.org 2022-06
Main Authors: Yao Rong, Leemann, Tobias, Borisov, Vadim, Kasneci, Gjergji, Kasneci, Enkelejda
Format: Article
Language:English
Subjects:
Evaluation
Information theory
Perturbation
Pixels
Quality assessment
Source code
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:With a variety of local feature attribution methods being proposed in recent years, follow-up work suggested several evaluation strategies. To assess the attribution quality across different attribution techniques, the most popular among these evaluation strategies in the image domain use pixel perturbations. However, recent advances discovered that different evaluation strategies produce conflicting rankings of attribution methods and can be prohibitively expensive to compute. In this work, we present an information-theoretic analysis of evaluation strategies based on pixel perturbations. Our findings reveal that the results are strongly affected by information leakage through the shape of the removed pixels as opposed to their actual values. Using our theoretical insights, we propose a novel evaluation framework termed Remove and Debias (ROAD) which offers two contributions: First, it mitigates the impact of the confounders, which entails higher consistency among evaluation strategies. Second, ROAD does not require the computationally expensive retraining step and saves up to 99% in computational costs compared to the state-of-the-art. We release our source code at https://github.com/tleemann/road_evaluation.
ISSN:2331-8422