Sanity Checks for Explanation Uncertainty

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Valdenegro-Toro, Matias, Mulye, Mihir
Formato: Preprint
Publicado: 2024
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866911813180850176
author Valdenegro-Toro, Matias
Mulye, Mihir
author_facet Valdenegro-Toro, Matias
Mulye, Mihir
contents Explanations for machine learning models can be hard to interpret or be wrong. Combining an explanation method with an uncertainty estimation method produces explanation uncertainty. Evaluating explanation uncertainty is difficult. In this paper we propose sanity checks for uncertainty explanation methods, where a weight and data randomization tests are defined for explanations with uncertainty, allowing for quick tests to combinations of uncertainty and explanation methods. We experimentally show the validity and effectiveness of these tests on the CIFAR10 and California Housing datasets, noting that Ensembles seem to consistently pass both tests with Guided Backpropagation, Integrated Gradients, and LIME explanations.
format Preprint
id arxiv_https___arxiv_org_abs_2403_17212
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Sanity Checks for Explanation Uncertainty
Valdenegro-Toro, Matias
Mulye, Mihir
Machine Learning
Artificial Intelligence
Explanations for machine learning models can be hard to interpret or be wrong. Combining an explanation method with an uncertainty estimation method produces explanation uncertainty. Evaluating explanation uncertainty is difficult. In this paper we propose sanity checks for uncertainty explanation methods, where a weight and data randomization tests are defined for explanations with uncertainty, allowing for quick tests to combinations of uncertainty and explanation methods. We experimentally show the validity and effectiveness of these tests on the CIFAR10 and California Housing datasets, noting that Ensembles seem to consistently pass both tests with Guided Backpropagation, Integrated Gradients, and LIME explanations.
title Sanity Checks for Explanation Uncertainty
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2403.17212