Search schemes for random optimization algorithms that preserve the asymptotic distribution
Loading...
External sources
External sources
Date
Advisor
Coadvisor
Graduate program
Undergraduate course
Journal Title
Journal ISSN
Volume Title
Publisher
Cambridge University Press (CUP)
Type
Article
Access right
Acesso aberto
External sources
External sources
Abstract
Markovian algorithms for estimating the global maximum or minimum of real valued functions defined on some domain Ω ⊂ ℝd are presented. Conditions on the search schemes that preserve the asymptotic distribution are derived. Global and local search schemes satisfying these conditions are analysed and shown to yield sharper confidence intervals when compared to the i.i.d. case. © 1999 Applied Probability Trust.
Description
Language
English
Citation
Journal of Applied Probability, v. 36, n. 3, p. 825-836, 1999.
