Optimal learning and experimentation in bandit problems

Journal of Economic Dynamics and Control

Volumn 27, Issue 1, 2002, Pages 87-108

  Brezzi, Monica ^a   Lai, Tze Leung ^b  

^a Ministero del Tesoro   (Italy)

^b STANFORD UNIVERSITY   (United States)

Author keywords

Corrected binomial algorithm; Incomplete learning; Multi armed bandits; Optimal stopping; Switching costs

Indexed keywords

EID: 0036334330     PISSN: 01651889     EISSN: None     Source Type: Journal    
DOI: 10.1016/S0165-1889(01)00028-8     Document Type: Article

References (24)

1. Asymptotically efficient adaptive allocation rules for the multiarmed bandit problem with switching costs
2. Denumerable-narmed bandits
3. Switching costs and the Gittins index
4. A Bernoulli two-armed bandit
5. Optimal stopping for Brownian motion in bandit problems and sequential analysis
6. Incomplete learning from endogenous data in dynamic allocation
7. Optimal stopping and dynamic allocation
8. Numerical solutions for Bayes sequential decision problems
9. Some remarks on the two-armed bandit
10. Contributions to the two-armed bandit problem
11. Bandit processes and dynamic allocation indices
12. Multi-Armed Bandit Allocation Indices
13. A dynamic allocation index for the sequential design of experiments
14. Job-search and the theory of turnover
15. A survey of some results in stochastic adaptive control
16. Adaptive treatment allocation and the multi-armed bandit problem
17. Asymptotically efficient adaptive allocation rules
18. On some asymptotic properties of maximum likelihood estimates and related Bayes estimates
19. Price disperson and incomplete learning in the long run
20. Job-search and labor market analysis
21. Sequential tests involving two populations
22. A two-armed bandit theory of market pricing
23. Multi-armed bandits and the Gittins index
24. Asymptotically optimal sequential tests of linear hypotheses in multiparameter exponential families