Tight Auditing of Differential Privacy in MST and AIM
Fuente:
arXiv
Salvato in:
| Autori principali: | , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866914492081766400 |
|---|---|
| author | Ganev, Georgi Annamalai, Meenatchi Sundaram Muthu Selva Kulynych, Bogdan |
| author_facet | Ganev, Georgi Annamalai, Meenatchi Sundaram Muthu Selva Kulynych, Bogdan |
| contents | State-of-the-art Differentially Private (DP) synthetic data generators such as MST and AIM are widely used, yet tightly auditing their privacy guarantees remains challenging. We introduce a Gaussian Differential Privacy (GDP)-based auditing framework that measures privacy via the full false-positive/false-negative tradeoff. Applied to MST and AIM under worst-case settings, our method provides the first tight audits in the strong-privacy regime. For $(ε,δ)=(1,10^{-2})$, we obtain $μ_{emp}\approx0.43$ vs. implied $μ=0.45$, showing a small theory-practice gap.
Our code is publicly available: https://github.com/sassoftware/dpmm. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_18352 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Tight Auditing of Differential Privacy in MST and AIM Ganev, Georgi Annamalai, Meenatchi Sundaram Muthu Selva Kulynych, Bogdan Cryptography and Security Artificial Intelligence Machine Learning State-of-the-art Differentially Private (DP) synthetic data generators such as MST and AIM are widely used, yet tightly auditing their privacy guarantees remains challenging. We introduce a Gaussian Differential Privacy (GDP)-based auditing framework that measures privacy via the full false-positive/false-negative tradeoff. Applied to MST and AIM under worst-case settings, our method provides the first tight audits in the strong-privacy regime. For $(ε,δ)=(1,10^{-2})$, we obtain $μ_{emp}\approx0.43$ vs. implied $μ=0.45$, showing a small theory-practice gap. Our code is publicly available: https://github.com/sassoftware/dpmm. |
| title | Tight Auditing of Differential Privacy in MST and AIM |
| topic | Cryptography and Security Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2604.18352 |