scientific article; zbMATH DE number 3891133

From MaRDI portal
Publication:3347666





[For the entire collection see Zbl 0553.00012.] A discrete Markov decision process with a finite state and action space in which there is multiplicity of expected average cost criteria is considered. The optimal solution of this problem is based on the concept of efficient policies. The existence of efficient policies in a subset of Markov policies is shown. Efficient policies are characterized as average optimal policies with respect to an appropriate single objective average cost criterion. The theoretical and practical possibilities of exploring this result are given.











This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3347666)