October 13, 2025Open Access

Adaptive Stochastic Gradient Descent Ascent Algorithm for Nonconvex Minimax Problems with Decision-Dependent Distributions

View Full Paper

Ask AI

Bookmark

ASGDA achieves $ε$-stationary point within $ ext{O}(ε^{-(4+δ)})$ evaluations, optimizing stochastic minimax objectives.
When satisfying Polyak-Łojasiewicz inequality, AASGDA finds $ε$-stationary points in $ ext{O}(κ_y^4ε^{-4})$ evaluations.
Algorithms dynamically learn distribution maps while solving nonconvex minimax problems effectively with real data.
Numerical experiments validate the effectiveness of both ASGDA and AASGDA in various settings.

Discussion

No takes yet. Share an insight, caveat, or question.

Proposed algorithms optimize nonconvex minimax problems with decision-dependent distributions, suggesting improved efficacy.

Gao et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark