Mutually Dependent Markov Decision Processes

抄録

In this paper, we introduce a basic framework for mutually dependent Markov decision processes (MDMDP) showing recursive mutual dependence. Our model is structured upon two types of finite-stage Markov decision processes. At each stage, the reward in one process is given by the optimal value of the alternative process problem, whose initial state is determined by the current state and decision in the original process. We formulate the MDMDP model and derive mutually dependent recursive equations by dynamic programming. Furthermore, MDMDP is illustrated in a numerical example. The model enables easier treatment of some classes of complex multi-stage decision processes.