Displaying similar documents to “Partially observable Markov decision processes with partially observable random discount factors”

Estimates of stability of Markov control processes with unbounded costs

Evgueni I. Gordienko, Francisco Salem-Silva (2000)

Kybernetika

Similarity:

For a discrete-time Markov control process with the transition probability p , we compare the total discounted costs V β ( π β ) and V β ( π ˜ β ) , when applying the optimal control policy π β and its approximation π ˜ β . The policy π ˜ β is optimal for an approximating process with the transition probability p ˜ . A cost per stage for considered processes can be unbounded. Under certain ergodicity assumptions we establish the upper bound for the relative stability index [ V β ( π ˜ β ) - V β ( π β ) ] / V β ( π β ) . This bound does not depend...

Deterministic optimal policies for Markov control processes with pathwise constraints

Armando F. Mendoza-Pérez, Onésimo Hernández-Lerma (2012)

Applicationes Mathematicae

Similarity:

This paper deals with discrete-time Markov control processes in Borel spaces with unbounded rewards. Under suitable hypotheses, we show that a randomized stationary policy is optimal for a certain expected constrained problem (ECP) if and only if it is optimal for the corresponding pathwise constrained problem (pathwise CP). Moreover, we show that a certain parametric family of unconstrained optimality equations yields convergence properties that lead to an approximation scheme which...

Adaptive control for discrete-time Markov processes with unbounded costs: Discounted criterion

Evgueni I. Gordienko, J. Adolfo Minjárez-Sosa (1998)

Kybernetika

Similarity:

We study the adaptive control problem for discrete-time Markov control processes with Borel state and action spaces and possibly unbounded one-stage costs. The processes are given by recurrent equations x t + 1 = F ( x t , a t , ξ t ) , t = 0 , 1 , ... with i.i.d. ℜ k -valued random vectors ξ t whose density ρ is unknown. Assuming observability of ξ t we propose the procedure of statistical estimation of ρ that allows us to prove discounted asymptotic optimality of two types of adaptive policies used early for the processes with bounded...