On compactness of the space of policies in stochastic dynamic programming

On compactness of the space of policies in stochastic dynamic programming

0.00 Avg rating0 Votes
Article ID: iaor19891083
Country: Netherlands
Volume: 32
Issue: 1
Start Page Number: 141
End Page Number: 150
Publication Date: Jun 1989
Journal: Stochastic Processes and Their Applications
Authors:
Abstract:

For stochastic dynamic programming models with compact, history-dependent action sets, sufficient conditions are given for the existence of a topology such that (i) the space of policies is compact, and (ii) the expected reward functionals are upper semicontinuous. The sufficient conditions (C) used here are weaker than both conditions (W) and (S) of Schäl, and unify these.

Reviews

Required fields are marked *. Your email address will not be published.