文档下载 文档下载导航
设为首页 | 加入收藏
搜索 请输入内容:  
 导航当前位置: 文档下载 > 所有分类 > Global optimisation Evolutionary neural networks Evolutionary algorithms Genetic algorithms
侵权投诉

Global optimisation Evolutionary neural networks Evolutionary algorithms Genetic algorithms

developments and future directions of bio-inspired

ECOLOGICALINFORMATICS1(2006)9–2

Global optimisation Evolutionary neural networks Evolutionary algorithms Genetic algorithms

2

availableatdoc.guandang.net

Global optimisation Evolutionary neural networks Evolutionary algorithms Genetic algorithms

m

doc.guandang.net

Currentdevelopmentsandfuturedirectionsofbio-inspiredcomputationandimplicationsforecoinformatics

XinYao*,YongLiu,JinLi,JunHe,ColinFrayn

TheCentreofExcellenceforResearchinComputationalIntelligenceandApplications(CERCIA),SchoolofComputerScience,TheUniversityofBirmingham,Edgabston,BirminghamB152TT,UK

ARTICLEINFO

Articlehistory:Received15July2005Accepted15July2005Keywords:

EvolutionarycomputationNeuralnetworkensemblesGlobaloptimisation

EvolutionaryneuralnetworksEvolutionaryalgorithmsGeneticalgorithms

ABSTRACT

Evolutionaryandneuralcomputationhasbeenusedwidelyinsolvingvariousproblemsinbiologicalecosystems.Thispaperreviewssomeoftherecentworkinevolutionarycomputationandneuralnetworkensemblesthatcouldbeexploredfurtherinthecontextofecoinformatics.Althoughthesebio-inspiredtechniqueswerenotdevelopedspecificallyforecoinformatics,theirsuccessesinsolvingcomplexproblemsinotherfieldsdemonstratehowthesetechniquescouldbeadaptedandusedfortacklingdifficultproblemsinecoinformatics.Firstly,wewillreviewourworkinmodellingandmodelcalibration,whichisanimportanttopicinecoinformatics.Secondlyoneexamplewillbegiventoillustratehowcoevolutionaryalgorithmscouldbeusedinproblem-solving.Thirdly,wewilldescribeourworkonneuralnetworkensembles,whichcanbeusedforvariousclassificationandpredictionproblemsinecoinformatics.Finally,wewilldiscussecosystem-inspiredcomputationalmodelsandalgorithmsthatcouldbeexploredasdirectionsoffutureresearch.

©2005ElsevierB.V.Allrightsreserved.

1.Introduction

Manyalgorithmsinevolutionarycomputationandneuralnet-workshavebeenusedinecoinformatics.However,therehavesomerecentnewdevelopmentsthatcouldbeveryusefulforsolvingcomplexecoinformaticsproblems.Thispaperreviewsandsummarisessomeoftherecentworkinevolutionarycomputationandneuralnetworkensemblesthatareparticu-larlyrelevanttoecoinformatics.

Data-drivenmodellingisamajortopicofresearchinecoin-formatics,wherethebestmodelistobefoundgivenasetofempiricaldata.Althoughitisnotnew,data-drivenmodellingisstillanunsolvedproblemespeciallyfornoisydata.Therearestatisticalandneuralnetworkapproachesthatcouldpotentiallyhelpinmodelling.However,oneofthedraw-backsforapproacheslikeneuralnetworksisthelackof

goodcomprehensibilityoftheresults.Itisoftenveryhardtounderstandandinterprettheneuralnetworkresults,un-lessadditionaltechniquessuchasruleextractionwereused.Evolutionarycomputationtechniques,e.g.,geneticprogram-ming,offeropportunitiestoevolvemodelstructuresaswellasmodelparametersexplicitly.Theirresultsaremorecom-prehensibleandeasiertointerpretthanthoseresultsgener-atedbytheblackboxmethods(e.g.,neuralnetworks).Wewillgiveoneexampleinastrophysicstoillustratehowdata-drivenmodellingcouldbedonebyatwostageevolu-tionaryapproach(Lietal.,2004).

Coevolutionisatypicalphenomenoninbiologicalecosys-tems.Itsideacanbeusedtodesignveryeffectiveproblem-solvingalgorithms.Thekeyideabehindcoevolutionaryalgo-rithmsisthatthefitnessofanindividualdependsonotherindividualsinthesamepopulation.Thesameindividualmay

Correspondingauthor.

E-mailaddress:x.yao@cs.bham.ac.uk(X.

Global optimisation Evolutionary neural networks Evolutionary algorithms Genetic algorithms

Yao).

1574-9541/$-seefrontmatter©2005ElsevierB.V.Allrightsreserved.doi:10.1016/j.ecoinf.2005.07.001

第1页

你可能喜欢

热门文档

相关文档

站点地图 | 文档上传 | 侵权投诉 | 手机版
本站所有资源均来自互联网,本站只负责收集和整理,均不承担任何法律责任,如有侵权等其它行为请联系我们.
文档下载 Copyright 2013 doc.guandang.net All Rights Reserved.  -  email