Consider a finite-state, finite-action Markovian Decision Process for which the state space has been partitioned into subsets. The decisionmaker can only observe the subset to which the states of the process belong, and not the actual states of the ...
