Search this article
Abstract
We introduce a new framework for dynamic programming called mutually dependent decision processes (MDDPs). Each MDDPs model is constructed from two or more finite-stage deterministic decision processes. At each stage, the reward in one process depends on the optimal values of the other processes, whose initial state is determined by the current state and decision of the original process. We formulate the MDDPs models and derive their mutually dependent recursive equations by dynamic programming.
Journal
-
- Bulletin of the Kyushu Institute of Technology. Pure and applied mathematics
-
Bulletin of the Kyushu Institute of Technology. Pure and applied mathematics (63), 15-26, 2016-03-31
Kyushu Institute of Technology Faculty of Engineering
- Tweet
Details 詳細情報について
-
- CRID
- 1390293556144424960
-
- NII Article ID
- 120005741790
-
- NII Book ID
- AA11237040
-
- ISSN
- 27589889
- 13438670
-
- HANDLE
- 10228/5616
-
- Text Lang
- en
-
- Data Source
-
- JaLC
- IRDB
- CiNii Articles
- KAKEN
-
- Abstract License Flag
- Allowed