A corrected and improved computational scheme for finite horizon partially observable Markov decision processes

A corrected and improved computational scheme for finite horizon partially observable Markov decision processes

0.00 Avg rating0 Votes
Article ID: iaor19912069
Country: Canada
Volume: 29
Issue: 3
Start Page Number: 206
End Page Number: 212
Publication Date: Aug 1991
Journal: INFOR
Authors: ,
Keywords: computational analysis, control processes
Abstract:

This paper corrects the algorithm given by Smallwood and Sondik for calculating optimal policies for partially observable Markov processes. It then develops an efficient algorithm for finding the optimal policy.

Reviews

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 Your email address will not be published.