OCGTMAMLDec 1, 2014

Game-theoretical control with continuous action sets

arXiv:1412.0543v13 citations
Originality Incremental advance
AI Analysis

This addresses control problems in distributed systems for applications like robotics or networks, but it is incremental as it extends existing methods to continuous settings.

The paper tackles the problem of designing distributed control systems using game-theoretical learning with continuous action sets, proposing an actor-critic reinforcement learning algorithm that provably converges to equilibrium in potential games.

Motivated by the recent applications of game-theoretical learning techniques to the design of distributed control systems, we study a class of control problems that can be formulated as potential games with continuous action sets, and we propose an actor-critic reinforcement learning algorithm that provably converges to equilibrium in this class of problems. The method employed is to analyse the learning process under study through a mean-field dynamical system that evolves in an infinite-dimensional function space (the space of probability distributions over the players' continuous controls). To do so, we extend the theory of finite-dimensional two-timescale stochastic approximation to an infinite-dimensional, Banach space setting, and we prove that the continuous dynamics of the process converge to equilibrium in the case of potential games. These results combine to give a provably-convergent learning algorithm in which players do not need to keep track of the controls selected by the other agents.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes