Note—A Counterexample in Continuous Markov Decision Chains

Abstract

Results in another paper in this issue may raise the question of whether for finite state and action, continuous time parameter Markov decision chains there always exists an initially stationary optimal policy. We give an example in which there is do such policy.