|
| << vorige | volgende >> |
Tijdschrift beschrijving
Alle jaargangen van het bijbehorende tijdschrift
Alle afleveringen van het bijbehorende jaargang
Alle artikelen van de bijbehorende aflevering
Details van artikel 10 van 19 gevonden artikelen
| Titel: |
Generalized two-stage bandit problem |
| Auteur: |
Yue, Jack C. |
| Verschenen in: |
Communications in statistics |
| Paginering: | Jaargang 28 (1999) nr. 9 pagina's 2261-2276 |
| Jaar: |
1999 |
| Inhoud: |
Two treatments which yield dichotomous outcomes are available for use in a clinical trial with a two-stage setting. Treatments are chosen sequentially in the first stage, and a single treatment (usually the better one) must be used in the second stage. The goal is to find the optimal strategy which maximizes the expected number of successes. The optimal strategy is to choose the unknown treatment when one treatment is known and when the number of patients is known, or unknown but satisfies certain regular conditions. In this paper, we extend the previous study by assuming that both treatments are unknown but that one treatment is better known (i.e. with smaller prior variance), and explore the conditions in which the better known treatment can be omitted in the first stage. |
| Uitgever: |
Taylor & Francis |
| Bronbestand: |
Elektronische Wetenschappelijke Tijdschriften |
Details van artikel 10 van 19 gevonden artikelen
| << vorige | volgende >> |