Time-varying Markov decision processes with state-action-dependent discount factors and unbounded costs

Title:

Time-varying Markov decision processes with state-action-dependent discount factors and unbounded costs

Creator:

Escobedo-Trujillo , Beatris A. and Higuera-Chan, Carmen G.

Identifier:

https://cdk.lib.cas.cz/client/handle/uuid:70f9b4bf-88fc-41c0-bd90-b9ebc75d6a48
uuid:70f9b4bf-88fc-41c0-bd90-b9ebc75d6a48
doi:10.14736/kyb-2019-1-0166

Subject:

discounted optimality, non-constant discount factor, and time-varying Markov decision processes

Type:

model:article and TEXT

Format:

bez média and svazek

Description:

In this paper we are concerned with a class of time-varying discounted Markov decision models Mn with unbounded costs cn and state-action dependent discount factors. Specifically we study controlled systems whose state process evolves according to the equation xn+1=Gn(xn,an,ξn),n=0,1,…, with state-action dependent discount factors of the form αn(xn,an), where an and ξn are the control and the random disturbance at time n, respectively. Assuming that the sequences of functions {αn},{cn} and {Gn} converge, in certain sense, to α∞, c∞ and G∞, our objective is to introduce a suitable control model for this class of systems and then, to show the existence of optimal policies for the limit system M∞ corresponding to α∞, c∞ and G∞. Finally, we illustrate our results and their applicability in a class of semi-Markov control models.

Language:

English

Rights:

http://creativecommons.org/publicdomain/mark/1.0/
policy:public

Coverage:

166-182

Source:

Kybernetika | 2019 Volume:55 | Number:1

Harvested from:

CDK

Metadata only:

false