Article ID: | iaor1994314 |
Country: | India |
Volume: | 14 |
Issue: | 2 |
Start Page Number: | 203 |
End Page Number: | 219 |
Publication Date: | May 1993 |
Journal: | Journal of Information & Optimization Sciences |
Authors: | White D.J. |
Keywords: | markov processes, decision theory, programming: dynamic, military & defence |
An heuristic procedure is presented for multi-dimensional Markov decision processes where current approximating optimal procedures are computationally demanding. It is applicable when certain simple policies are easy to evaluate, and it is not necessary to evaluate the improved policies. Bounds on the loss of optimality arising from such policies are given, which are used to reject or accept any policy derived.