An adaptive niching EDA with balance searching based on clustering analysis

Benhui Chen, Takayuki Furuzuki

Research output: Contribution to journalArticle

Abstract

For optimization problems with irregular and complex multimodal landscapes, Estimation of Distribution Algorithms (EDAs) suffer from the drawback of premature convergence similar to other evolutionary algorithms. In this paper, we propose an adaptive niching EDA based on Affinity Propagation (AP) clustering analysis. The AP clustering is used to adaptively partition the niches and mine the searching information from the evolution process. The obtained information is successfully utilized to improve the EDA performance by using a balance niching searching strategy. Two different categories of optimization problems are used to evaluate the proposed adaptive niching EDA. The first one is solving three benchmark functional multimodal optimization problems by a continuous EDA based on single Gaussian probabilistic model; the other one is solving a real, complicated discrete EDA optimization problem, the HP model protein folding based on k-order Markov probabilistic model. Simulation results show that the proposed adaptive niching EDA is an efficient method.

Original languageEnglish
Pages (from-to)1792-1799
Number of pages8
JournalIEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences
VolumeE93-A
Issue number10
DOIs
Publication statusPublished - 2010 Oct

Fingerprint

Niching
Clustering Analysis
Optimization Problem
Probabilistic Model
Affine transformation
Propagation
Partitions (building)
Protein folding
Multimodal Optimization
Protein Folding
Premature Convergence
Niche
Gaussian Model
Evolutionary algorithms
Markov Model
Evolutionary Algorithms
Irregular
Partition
Clustering
Benchmark

Keywords

  • Affinity propagation (ap) clustering
  • Boltzmann scheme
  • Estimation of distribution algorithms (edas)
  • Niching clearing procedure

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Computer Graphics and Computer-Aided Design
  • Applied Mathematics
  • Signal Processing

Cite this

@article{e99e2caeb50442899344090e157fe672,
title = "An adaptive niching EDA with balance searching based on clustering analysis",
abstract = "For optimization problems with irregular and complex multimodal landscapes, Estimation of Distribution Algorithms (EDAs) suffer from the drawback of premature convergence similar to other evolutionary algorithms. In this paper, we propose an adaptive niching EDA based on Affinity Propagation (AP) clustering analysis. The AP clustering is used to adaptively partition the niches and mine the searching information from the evolution process. The obtained information is successfully utilized to improve the EDA performance by using a balance niching searching strategy. Two different categories of optimization problems are used to evaluate the proposed adaptive niching EDA. The first one is solving three benchmark functional multimodal optimization problems by a continuous EDA based on single Gaussian probabilistic model; the other one is solving a real, complicated discrete EDA optimization problem, the HP model protein folding based on k-order Markov probabilistic model. Simulation results show that the proposed adaptive niching EDA is an efficient method.",
keywords = "Affinity propagation (ap) clustering, Boltzmann scheme, Estimation of distribution algorithms (edas), Niching clearing procedure",
author = "Benhui Chen and Takayuki Furuzuki",
year = "2010",
month = "10",
doi = "10.1587/transfun.E93.A.1792",
language = "English",
volume = "E93-A",
pages = "1792--1799",
journal = "IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences",
issn = "0916-8508",
publisher = "Maruzen Co., Ltd/Maruzen Kabushikikaisha",
number = "10",

}

TY - JOUR

T1 - An adaptive niching EDA with balance searching based on clustering analysis

AU - Chen, Benhui

AU - Furuzuki, Takayuki

PY - 2010/10

Y1 - 2010/10

N2 - For optimization problems with irregular and complex multimodal landscapes, Estimation of Distribution Algorithms (EDAs) suffer from the drawback of premature convergence similar to other evolutionary algorithms. In this paper, we propose an adaptive niching EDA based on Affinity Propagation (AP) clustering analysis. The AP clustering is used to adaptively partition the niches and mine the searching information from the evolution process. The obtained information is successfully utilized to improve the EDA performance by using a balance niching searching strategy. Two different categories of optimization problems are used to evaluate the proposed adaptive niching EDA. The first one is solving three benchmark functional multimodal optimization problems by a continuous EDA based on single Gaussian probabilistic model; the other one is solving a real, complicated discrete EDA optimization problem, the HP model protein folding based on k-order Markov probabilistic model. Simulation results show that the proposed adaptive niching EDA is an efficient method.

AB - For optimization problems with irregular and complex multimodal landscapes, Estimation of Distribution Algorithms (EDAs) suffer from the drawback of premature convergence similar to other evolutionary algorithms. In this paper, we propose an adaptive niching EDA based on Affinity Propagation (AP) clustering analysis. The AP clustering is used to adaptively partition the niches and mine the searching information from the evolution process. The obtained information is successfully utilized to improve the EDA performance by using a balance niching searching strategy. Two different categories of optimization problems are used to evaluate the proposed adaptive niching EDA. The first one is solving three benchmark functional multimodal optimization problems by a continuous EDA based on single Gaussian probabilistic model; the other one is solving a real, complicated discrete EDA optimization problem, the HP model protein folding based on k-order Markov probabilistic model. Simulation results show that the proposed adaptive niching EDA is an efficient method.

KW - Affinity propagation (ap) clustering

KW - Boltzmann scheme

KW - Estimation of distribution algorithms (edas)

KW - Niching clearing procedure

UR - http://www.scopus.com/inward/record.url?scp=77957805418&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77957805418&partnerID=8YFLogxK

U2 - 10.1587/transfun.E93.A.1792

DO - 10.1587/transfun.E93.A.1792

M3 - Article

AN - SCOPUS:77957805418

VL - E93-A

SP - 1792

EP - 1799

JO - IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences

JF - IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences

SN - 0916-8508

IS - 10

ER -