Paper Info
Title
Interaction-Detection Metric With Differential Mutual Complement For Dependency Structure Matrix Genetic Algorithm
Abstract
Dependency structure matrix genetic algorithm (DSMGA), one of estimation of distribution algorithms (EDAs), builds models via dependency structure matrix clustering techniques. Previous researches have shown that DSMGA can effectively solve nearly decomposable problems. The efficiency of DSMGA and other model-building GAs greatly depend on their interaction-detection metrics. This paper investigates three commonly used metrics, nonlinearity, simultaneity, and entropy. Then it proposes a new interaction-detection metric which aims at what GAs really need. The proposed metric, namely the differential mutual complement, is based on both the disruption and reproduction effects of the crossover operator on significant schemata. Empirical results show that DSMGA with the proposed metric performs better than the other existing metrics on the aspects of sensitivity to threshold and function evaluation. This new metric is shown to perform well on DSMGA and could be expected to work with other EDAs.
Year
DOI
Venue
2010
10.1109/CEC.2010.5586098
2010 IEEE CONGRESS ON EVOLUTIONARY COMPUTATION (CEC)
Keywords
Field
DocType
entropy,sensitivity,genetic algorithm,mutual information,estimation of distribution algorithm,model building,genetic algorithms,dependency structure matrix,mathematical model
EDAS,Mathematical optimization,Crossover,Estimation of distribution algorithm,Computer science,Artificial intelligence,Mutual information,Design structure matrix,Cluster analysis,Genetic algorithm,Machine learning,Simultaneity
Conference
Citations 
PageRank 
References 
5
0.44
7
Authors
4
Name
Order
Citations
PageRank
Kai-Chun Fan1121.58
Jui-Ting Lee2131.92
Tian-Li Yu343035.28
Tsung-Yu Ho4162.78