Convex-Concave Zero-Sum Markov Stackelberg Games

Denizalp Goktas, Arjun Prakash, Amy Greenwald

Keywords: Stackelberg games, Equilibrium Computation, Policy Gradient

TL;DR: We develop policy gradient methods to solve zero-sum stochastic Stackelberg games from observed play, and apply our methods to solve reach avoid problems.

Abstract: Zero-sum Markov Stackelberg games can be used to model myriad problems, in domains ranging from economics to human robot interaction. In this paper, we develop policy gradient methods that solve these games in continuous state and action settings using noisy gradient estimates computed from observed trajectories of play. When the games are convex-concave, we prove that our algorithms converge to Stackelberg equilibrium in polynomial time. We also show that reach-avoid problems are naturally modeled as convex-concave zero-sum Markov Stackelberg games, and that Stackelberg equilibrium policies are more effective than their Nash counterparts in these problems.

Submission Number: 15332