Paper ID: 2302.04972

Differentially Private Optimization for Smooth Nonconvex ERM

Changyu Gao, Stephen J. Wright

We develop simple differentially private optimization algorithms that move along directions of (expected) descent to find an approximate second-order solution for nonconvex ERM. We use line search, mini-batching, and a two-phase strategy to improve the speed and practicality of the algorithm. Numerical experiments demonstrate the effectiveness of these approaches.

Submitted: Feb 9, 2023