Benutzer: Gast  Login
Titel:

Gaussian Membership Inference Privacy

Dokumenttyp:
Konferenzbeitrag
Art des Konferenzbeitrags:
Textbeitrag / Aufsatz
Autor(en):
Leemann, Tobias; Pawelczyk, Martin; Kasneci, Gjergji
Abstract:
We propose a novel and practical privacy notion called f-Membership Inference Privacy (f-MIP), which explicitly considers the capabilities of realistic adversaries under the membership inference attack threat model. Consequently, f-MIP offers interpretable privacy guarantees and improved utility (e.g., better classification accuracy). In particular, we derive a parametric family of f-MIP guarantees that we refer to as μ-Gaussian Membership Inference Privacy (μ-GMIP) by theoretically analyzing...     »
Kongress- / Buchtitel:
Advances in Neural Information Processing Systems (NeurIPS)
Ausgabe:
36
Jahr:
2023
Reviewed:
ja
WWW:
https://proceedings.neurips.cc/paper_files/paper/2023/hash/e9df36b21ff4ee211a8b71ee8b7e9f57-Abstract-Conference.html
 BibTeX