Surrogate Model Selection for Evolutionary Multiobjective Optimization

被引：0

作者：

Pilat, Martin ^{[1
]}

Neruda, Roman ^{[2
]}

机构：

[1] Charles Univ Prague, Fac Math & Phys, Prague, Czech Republic

[2] Acad Sci Czech Republ, Inst Comp Sci, Prague 8, Czech Republic

来源：

2013 IEEE CONGRESS ON EVOLUTIONARY COMPUTATION (CEC) | 2013年

关键词：

Multiobjective optimization; meta-model; evolutionary algorithm; model selection;

D O I：

暂无

中图分类号：

TP301 [理论、方法];

学科分类号：

081202 ;

摘要：

In surrogate evolutionary algorithms, usually the type of surrogate model is chosen beforehand, and it is never changed during the run of the evolution. Moreover, the reasoning why a particular type of model was chosen is often missing. In this paper, we present a framework which in each generation selects the most suitable surrogate from a set of models based on some pre-defined criteria. The results based on different types of model selectors are compared, and the dynamics of the evolution together with the change of the selected model type during the run of the evolutionary algorithm are discussed.

引用

页码：1860 / 1867

页数：8

共 17 条

[1]

[Anonymous], 2008, Proc. of 2008 IEEE Congress on Evolutionary Computation, DOI DOI 10.1109/CEC.2008.4631121

[2]

Deb K., 2000, Parallel Problem Solving from Nature PPSN VI. 6th International Conference. Proceedings (Lecture Notes in Computer Science Vol.1917), P849

[3]

Deb K., 1995, Complex Systems, V9, P115

[4]

Deb K., 1996, Computer Science and informatics, V26, P30

[5]

DIAZMANRIQUEZ A, 2011, EV COMP CEC 2011 IEE, P2155

[6]

Hall M., 2009, SIGKDD Explorations, V11, P10, DOI DOI 10.1145/1656274.1656278

[7] Surrogate-assisted evolutionary computation: Recent advances and future challenges [J].