Online Stackelberg Optimization via Nonlinear Control

arXiv:2406.18805v1 Announce Type: new
Abstract: In repeated interaction problems with adaptive agents, our objective often requires anticipating and optimizing over the space of possible agent responses. We show that many problems of this form can be cast as instances of online (nonlinear) control which satisfy \textit{local controllability}, with convex losses over a bounded state space which encodes agent behavior, and we introduce a unified algorithmic framework for tractable regret minimization in such cases. When the instance dynamics are known but …

abstract agent agents arxiv cast control cs.gt cs.lg form instances losses optimization responses show space state type via