Publication Details
Bayesovské optimalizační algoritmy v dynamickém prostředí
BOA, dynamic problems, evolutionary optimization
Thispaper is an experimental study investigating the capability ofBayesian optimization algorithms to solve dynamic problems. We testedthe performance of two types of Bayesian optimization algorithms -Mixed continuous-discrete Bayesian Optimization Algorithm (MBOA) [1],and AdaptiveMixed Bayesian Optimization Algorithm (AMBOA)[2].We have compared the behaviour of both algorithms on a simple dynamic environment defined as a time-varying function with predefinedparameters. The experimental results confirmed the capability of bothBOA algorithms to adapt the search process, but for a limitedenvironment change. The AMBOA with the variance adaptationoutperformed the MBOA algorithm.
@inproceedings{BUT22417,
author="Miloš {Kobliha}",
title="Bayesovské optimalizační algoritmy v dynamickém prostředí",
booktitle="Sborník příspevků ze semináře Počítačové Architektury & Diagnostika",
year="2005",
pages="25--30",
publisher="České vysoké učení technické",
address="Praha",
isbn="80-01-03298-1"
}