Explainable AI: Efficiency Sequential Shapley Updating Approach

Shapley value-based explainable AI has recently attracted significant interest. However, the computational complexity of the Shapley value grows exponentially with the number of players, resulting in high computational costs that prevent its widespread practical application. To address this challeng...

Full description

Saved in:
Bibliographic Details
Main Authors: Ovanes Petrosian, Jinying Zou
Format: Article
Language:English
Published: IEEE 2024-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10750003/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850162028980731904
author Ovanes Petrosian
Jinying Zou
author_facet Ovanes Petrosian
Jinying Zou
author_sort Ovanes Petrosian
collection DOAJ
description Shapley value-based explainable AI has recently attracted significant interest. However, the computational complexity of the Shapley value grows exponentially with the number of players, resulting in high computational costs that prevent its widespread practical application. To address this challenge, various approximation methods have been proposed in the literature for computing the Shapley value, such as linear Shapley computation, sampling-based Shapley computation, and several estimation-based approaches. Among these methods, the sampling approach exhibits non-zero bias and variance and is sufficiently universal to be used with almost any AI algorithm. However, it suffers from unstable interpretability results and slow convergence in high-dimensional problems. To address these problems, we propose integrating a sequential Bayesian updating framework into the Shapley sampling approach. The core idea of this method is to dynamically update probabilities based on each sample’s Shapley value combined with a selection strategy. Both theoretical analysis and empirical results show that this method significantly improves the convergence speed and interpretability compared to the original sampling approach.
format Article
id doaj-art-38aaced70b204015b77ffebd60f20670
institution OA Journals
issn 2169-3536
language English
publishDate 2024-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-38aaced70b204015b77ffebd60f206702025-08-20T02:22:40ZengIEEEIEEE Access2169-35362024-01-011216641416642310.1109/ACCESS.2024.349554310750003Explainable AI: Efficiency Sequential Shapley Updating ApproachOvanes Petrosian0Jinying Zou1https://orcid.org/0000-0003-4641-4017Faculty of Applied Mathematics and Control Processes, Saint Petersburg State University, Saint Petersburg, RussiaFaculty of Applied Mathematics and Control Processes, Saint Petersburg State University, Saint Petersburg, RussiaShapley value-based explainable AI has recently attracted significant interest. However, the computational complexity of the Shapley value grows exponentially with the number of players, resulting in high computational costs that prevent its widespread practical application. To address this challenge, various approximation methods have been proposed in the literature for computing the Shapley value, such as linear Shapley computation, sampling-based Shapley computation, and several estimation-based approaches. Among these methods, the sampling approach exhibits non-zero bias and variance and is sufficiently universal to be used with almost any AI algorithm. However, it suffers from unstable interpretability results and slow convergence in high-dimensional problems. To address these problems, we propose integrating a sequential Bayesian updating framework into the Shapley sampling approach. The core idea of this method is to dynamically update probabilities based on each sample’s Shapley value combined with a selection strategy. Both theoretical analysis and empirical results show that this method significantly improves the convergence speed and interpretability compared to the original sampling approach.https://ieeexplore.ieee.org/document/10750003/Explainable AIinterpretabilitysequential Shapley updatingShapley valuesampling methodBayesian updating
spellingShingle Ovanes Petrosian
Jinying Zou
Explainable AI: Efficiency Sequential Shapley Updating Approach
IEEE Access
Explainable AI
interpretability
sequential Shapley updating
Shapley value
sampling method
Bayesian updating
title Explainable AI: Efficiency Sequential Shapley Updating Approach
title_full Explainable AI: Efficiency Sequential Shapley Updating Approach
title_fullStr Explainable AI: Efficiency Sequential Shapley Updating Approach
title_full_unstemmed Explainable AI: Efficiency Sequential Shapley Updating Approach
title_short Explainable AI: Efficiency Sequential Shapley Updating Approach
title_sort explainable ai efficiency sequential shapley updating approach
topic Explainable AI
interpretability
sequential Shapley updating
Shapley value
sampling method
Bayesian updating
url https://ieeexplore.ieee.org/document/10750003/
work_keys_str_mv AT ovanespetrosian explainableaiefficiencysequentialshapleyupdatingapproach
AT jinyingzou explainableaiefficiencysequentialshapleyupdatingapproach