Bernoulli two-armed bandits with geometric termination
The standard Bernoulli two-armed bandit model is modified by terminating the choice problem after the first unsuccessful trial. Both terminal reward situations and instances in which payoffs accrue with each success are considered. For independent machines, the stay-on-a-winner rule holds in each of these instances. Moreover, for the terminal payoff case, staying on a winner is optimal with interdependent machines. Increased prior information concerning the properties of a machine decreases its attractiveness by diminishing the prospect for long-term survival.
Year of publication: |
1981
|
---|---|
Authors: | Berry, Donald A. ; Viscusi, W. Kip |
Published in: |
Stochastic Processes and their Applications. - Elsevier, ISSN 0304-4149. - Vol. 11.1981, 1, p. 35-45
|
Publisher: |
Elsevier |
Keywords: | Two-armed bandit Bernoulli bandit stay-on-a-winner geometric termination |
Saved in:
Online Resource
Saved in favorites
Similar items by person
-
Discrete search with directional information
Berry, Donald A., (1986)
-
Detecting trends in arrangements of ordered objects : a likelihood approach
Berry, Donald A., (1979)
-
Berry, Scott M., (2004)
- More ...