Year: 2025
Type: article
Abstract: . In this work, we consider constrained stochastic optimization problems given hidden convexity, i.e., those that admit a convex reformulation via nonlinear (but invertible) map [Formula: see text]. A number of nonconvex problems ranging from optimal control, revenue, and inventory management to convex reinforcement learning all admit such a hidden convex structure. Unfortunately, in the majority of applications considered, the map [Formula: see text] is unavailable or implicit; therefore, directly solving the convex reformulation is not possible. On the other hand, the stochastic gradients with respect to the original variable are often easy to obtain. Motivated by these observations, we examine the basic projected stochastic (sub) gradient methods for solving such problems given hidden convexity. We provide the first sample complexity guarantees for global convergence in smooth and nonsmooth settings. Additionally, in the smooth setting, we improve our results to the last iterate convergence in terms of function value gap using the momentum variant of projected stochastic gradient descent.
Source: SIAM Journal on Optimization
DOI: 10.1137/22m1708903
Authors: Ilyas Fatkhullin, Niao He, Yifan Hu
Institutions: ETH Zurich, Rutgers, The State University of New Jersey
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W4416065663
