Feature Inference Attack on Shapley Values
Xinjian Luo, Yang-Fan Jiang, Xiaokui Xiao · Proceedings of the 2022 ACM SIGSAC Conference on Computer and Communications Security · 2022
As a solution concept in cooperative game theory, Shapley value is highly recognized in model interpretability studies and widely adopted by the leading Machine Learning as a Service (MLaaS) providers, such as Google, Microsoft, and IBM. However, as the Shapley value-based model interpretability methods have been thoroughly studied, few researchers consider the privacy risks incurred by Shapley values, despite that interpretability and privacy are two foundations of machine learning (ML) models.