all InfoSec News
Private Zeroth-Order Nonsmooth Nonconvex Optimization
July 1, 2024, 4:14 a.m. | Qinzi Zhang, Hoang Tran, Ashok Cutkosky
cs.CR updates on arXiv.org arxiv.org
Abstract: We introduce a new zeroth-order algorithm for private stochastic optimization on nonconvex and nonsmooth objectives. Given a dataset of size $M$, our algorithm ensures $(\alpha,\alpha\rho^2/2)$-R\'enyi differential privacy and finds a $(\delta,\epsilon)$-stationary point so long as $M=\tilde\Omega\left(\frac{d}{\delta\epsilon^3} + \frac{d^{3/2}}{\rho\delta\epsilon^2}\right)$. This matches the optimal complexity of its non-private zeroth-order analog. Notably, although the objective is not smooth, we have privacy ``for free'' whenever $\rho \ge \sqrt{d}\epsilon$.
algorithm alpha arxiv complexity cs.cr cs.lg dataset delta differential privacy epsilon math.oc non objectives omega optimization order point privacy private size
More from arxiv.org / cs.CR updates on arXiv.org
Jobs in InfoSec / Cybersecurity
Senior Network Architect - Wireless & Security
@ NVIDIA | US, CA, Santa Clara
Senior IT Auditor (Hybrid)
@ Progressive Leasing | Hybrid - Draper
Senior IT Auditor (Remote)
@ Progressive Leasing | Atlanta HUB
Consultant Directeur Audit Interne / Contrôle Interne | Assurance | CDI | H/F
@ PwC | Paris - Crystal Park
Principal Engineer - Network Security
@ Broadcom | USA-CA - Promontory E
Cryptologic Computer Scientist
@ Synergy ECP | Columbia, MD