STUDIES OF NEW APPLICATION AREAS OF BAYESIAN STOCHASTIC DYNAMIC PROGRAMMING FOR MANAGEMENT AND INFORMATIONAL SCIENCES
| Project/Area Number |
15510136
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Social systems engineering/Safety system
|
|---|
| Research Institution | UNIVERSITY OF HYOGO (2004)
Kobe University of Commerce (2003) |
|---|
Principal Investigator |
HAMADA Toshio UNIVERSITY OF HYOGO, SCHOOL OF BUSINESS ADMINISTRATION, PROFESSOR, 経営学部, 教授 (40116083)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
KIKUTA Kensaku UNIVERSITY OF HYOGO, SCHOOL OF BUSINESS ADMINISTRATION, PROFESSOR, 経営学部, 教授 (30126487)
KINIWA Jun UNIVERSITY OF HYOGO, SCHOOL OF ECONOMICS, PROFESSOR, 経営学部, 教授 (90177882)
|
|---|
| Project Period (FY) |
2003 – 2004
|
| Project Status |
Completed (Fiscal Year 2004)
|
| Budget Amount *help |
¥3,700,000 (Direct Cost: ¥3,700,000)
Fiscal Year 2004: ¥1,500,000 (Direct Cost: ¥1,500,000)
Fiscal Year 2003: ¥2,200,000 (Direct Cost: ¥2,200,000)
|
|---|
| Keywords | STOCHASIC DYNAMIC PROGRAMMING / SEQUENTIAL DECISION PROBLEM / BATCH SIZE / SCHEDULING / BAYES THEOREM / OPTIMAL STRATEGY / PAGING / GAME THEORY / 平均滞留時間 / ページ置換えアルゴリズム / First-In First-Out(FIFO) / Least Recently Used(LRU) / ランダムウォーク / マルコフ連鎖 / ヒット率 / 動的計画法 |
|---|
| Research Abstract |
In the area of informational science, the staying time of a page in the fast memory of computer is studied as a criterion and the result of this research will be presented in the IFORS international meeting. In the area of management and informational sciences, although several papers about the problems to decide the batch size and sequencing as the deterministic ones had been published in several journals, the problem had not been discussed under uncertainty in except one paper where only the decision of batch size is considered and the recursive equation is solved for the cases of job size less than or equal to 4. In our research project, the problem to decide the batch size is discussed for job size n≧4, where there are n jobs whose processing times are the same and known and the setup time is the random variable with an unknown parameter. This problem is analyzed from the Bayesian point of view, that is, the conjugate prior distribution for the unknown parameter is assumed to exist, the problem is formulated by dynamic programming and derived several properties of the objective function, and the optimal strategy is derived. The result of the research is presented in 2004 fall meeting of the Operations Research Society of Japan. Also, the problem to restrict the batch size as one or two is discussed and the result is presented in the meeting of the Mathematical Institute of Kyoto University. The problem to decide the batch size and sequencing simultaneously is also considered and several results are obtained and presented in 2005 spring meeting of the Operations Research Society of Japan. In the area of game and search theories, 3 papers are published in the related journals and a paper is presented in the international meeting in Arizona, USA.
|
Report
(3 results)
Research Products
(19 results)
