PANORAMIA: Privacy Auditing of Machine Learning Models without Retraining

arXiv:2402.09477v1 Announce Type: cross
Abstract: We introduce a privacy auditing scheme for ML models that relies on membership inference attacks using generated data as "non-members". This scheme, which we call PANORAMIA, quantifies the privacy leakage for large-scale ML models without control of the training process or model re-training and only requires access to a subset of the training data. To demonstrate its applicability, we evaluate our auditing scheme across multiple ML domains, ranging from image and tabular data classification to …

abstract arxiv attacks call control cs.cr cs.lg data generated inference machine machine learning machine learning models ml models privacy process retraining scale training type