Автор |
Hamada, Toshio |
Дата выпуска |
1992 |
dc.description |
The problem considered in this paper is to decide when to stop a sequence of the independent random variables from the uniform distribution on the interval (-1,u). Suppose that the true value of u is unknown and there is the prior knowledge that u has the Pareto distribution with parameters w and a as a prior distribution. The .objective is to maximize the total expected discounted sum of the observations. This problem is formulated by dynamic programming and the optimal strategy is denoted by the critical value function whose value for any parameter vector-value is easily calculated. |
Формат |
application.pdf |
Издатель |
Marcel Dekker, Inc. |
Копирайт |
Copyright Taylor and Francis Group, LLC |
Тема |
bandit problem |
Тема |
discount factor |
Тема |
Bayesian updatingt |
Тема |
optimal strategy |
Тема |
myopic strategy |
Название |
A doscounted uniform one-armed bandit problem |
Тип |
research-article |
DOI |
10.1080/07474949208836241 |
Electronic ISSN |
1532-4176 |
Print ISSN |
0747-4946 |
Журнал |
Sequential Analysis |
Том |
11 |
Первая страница |
1 |
Последняя страница |
15 |
Аффилиация |
Hamada, Toshio; Department of Management and Information Sciences, Himeji College of Hyogo |
Выпуск |
1 |
Библиографическая ссылка |
Beckmann, M. J. 1973. Der diskontierte Bandit. OR-Verfahren, : 9–18. XVIII |
Библиографическая ссылка |
Fischer, J. 1979. Der diskontierte Einarmige Bandit. Metrika, 26: 195–204. |
Библиографическая ссылка |
Fristedt, B. and Berry, D. A. 1988. Optimality of myopic stopping times for geometric discounting. J. Appl. Prob., 25: 473–443. |
Библиографическая ссылка |
Glazebrook, K. D. and Jones, D. M. 1983. Some best possible results for a discounted one arm bandit. Metrika, 30: 109–115. |
Библиографическая ссылка |
Hamada, T. 1985. Further results on a uniform two-armed bandit problem with one arm known. J. Japan Statist. Soc., 15: 193–208. |