Michael (Min Chen Chen)

Introduction As machine learning models become increasingly powerful and pervasive, concerns about privacy in their deployment have grown. One notable risk comes from membership inference attacks, where an adversary attempts to determine whether a particular data point was used to train a given model. For example, in healthcare, a membership inference attack could reveal if someone participated in a clinical study, threatening patient confidentiality. This risk motivates my work in this post, where I document my recent experiment investigating the possibility in using cumulative topological distance in membership inference attacks. ...