Papers
Topics
Authors
Recent
Search
2000 character limit reached

Private Zeroth-Order Nonsmooth Nonconvex Optimization

Published 27 Jun 2024 in math.OC, cs.CR, and cs.LG | (2406.19579v1)

Abstract: We introduce a new zeroth-order algorithm for private stochastic optimization on nonconvex and nonsmooth objectives. Given a dataset of size MM, our algorithm ensures (α,αρ2/2)(\alpha,\alpha\rho^2/2)-R\'enyi differential privacy and finds a (δ,ϵ)(\delta,\epsilon)-stationary point so long as M=Ω~(dδϵ3+d3/2ρδϵ2)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 ρ≥dϵ\rho \ge \sqrt{d}\epsilon.

Citations (4)

Summary

No one has generated a summary of this paper yet.

Paper to Video (Beta)

No one has generated a video about this paper yet.

Whiteboard

No one has generated a whiteboard explanation for this paper yet.

Open Problems

We haven't generated a list of open problems mentioned in this paper yet.

Continue Learning

We haven't generated follow-up questions for this paper yet.

Tweets

Sign up for free to view the 1 tweet with 0 likes about this paper.