This article presents a theoretical framework for probably approximately correct (PAC) multi-agent reinforcement learning (MARL) algorithms Markov games. Using the idea of delayed Q-learning, this extends well-known Nash Q-learning algorithm to build new PAC MARL general-sum In addition guiding design provably algorithm, enables checking whether an arbitrary is PAC. Comparative numerical result...

We consider a large family of problems in which an ordering (or, more precisely, chain subsets) finite set must be chosen to minimize some weighted sum costs. This includes variations min cover, several scheduling and search problems, Boolean function evaluation. define new problem, called the problem (MSOP), generalizes all these using cost weight defined on subsets set. Assuming polynomial ti...