Decentralized Bandit Learning of Markov Perfect Equilibria
Not on the map. Why?
Abstract (arXiv)
We study decentralized bandit learning in discounted infinite-horizon, general-sum Markov games with continuous actions. Agents observe the public state, but only their own realized rewards, and use simultaneous perturbation stochastic approximation to update their actions. Under strong monotonicity of the negative running-reward gradients and sufficiently small time steps, we prove that the fully decentralized procedure converges almost surely to the unique stationary Markov perfect equilibrium. We obtain an O(n-1/2) convergence rate and present a natural extension to episodic finite-horizon learning.
Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 15 of its 25 references that are in the map. First posted on arXiv 9 Oct 2026.
Builds on
15 of 25 references.
References · 15
- 2016Learning in games with continuous action sets and unknown payoff functionscited by 81
- 2021Decentralized Q-Learning in Zero-sum Markov Gamescited by 58
- 2015Decentralized Q-Learning for Stochastic Teams and Gamescited by 35
- 2018Bandit learning in concave N-person gamescited by 35
- 2021Independent Learning in Stochastic Gamescited by 22
- 2022Independent and Decentralized Learning in Markov Potential Gamescited by 22
- 2021Doubly Optimal No-Regret Online Learning in Strongly Monotone Games with Bandit Feedbackcited by 15
- 2019Decentralized Learning for Optimality in Stochastic Dynamic Teams and Games with Local Control and Global State Informationcited by 13
- 2021Improved Rates for Derivative Free Gradient Play in Strongly Monotone Gamescited by 7
- 2017Learning Generalized Nash Equilibria in a Class of Convex Gamescited by 4
- 2023Unsynchronized Decentralized Q-Learning: Two Timescale Analysis By Persistencecited by 4
- 2022Decentralized Policy Gradient for Nash Equilibria Learning of General-sum Stochastic Gamescited by 2
3 more
Built on by
Not cited yet.