This paper describes sufficient conditions for the existence of optimal policies for partially observable Markov decision processes (POMDPs) with Borel state, observation, and action sets, when the ...