Quantifying Membership Inference Vulnerability via Generalization Gap and Other Model Metrics
University of Central Florida（中佛罗里达大学）；University of Texas San Antonio（德克萨斯大学圣安东尼奥分校）
英文摘要：
We demonstrate how a target model's generalization gap leads directly to an effective deterministic black box membership inference attack (MIA). This provides an upper bound on how secure a model can be to MIA based on a simple metric. Moreover, this attack is shown to be optimal in the expected sense given access to only certain likely obtainable metrics regarding the network's training and performance. Experimentally, this attack is shown to be comparable in accuracy to state-of-art MIAs in many cases.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
