Transactions on Computational Science VIII

175 Pages·2011·6.466 MB·English

by  Manabu Gouko, Koji Ito (auth.), Marina L. Gavrilova, C. J. Kenneth Tan (eds.)

Lecture Notes in Computer Science 6260
CommencedPublicationin1973
FoundingandFormerSeriesEditors:
GerhardGoos,JurisHartmanis,andJanvanLeeuwen
EditorialBoard
DavidHutchison
LancasterUniversity,UK
TakeoKanade
CarnegieMellonUniversity,Pittsburgh,PA,USA
JosefKittler
UniversityofSurrey,Guildford,UK
JonM.Kleinberg
CornellUniversity,Ithaca,NY,USA
FriedemannMattern
ETHZurich,Switzerland
JohnC.Mitchell
StanfordUniversity,CA,USA
MoniNaor
WeizmannInstituteofScience,Rehovot,Israel
OscarNierstrasz
UniversityofBern,Switzerland
C.PanduRangan
IndianInstituteofTechnology,Madras,India
BernhardSteffen
TUDortmundUniversity,Germany
MadhuSudan
MicrosoftResearch,Cambridge,MA,USA
DemetriTerzopoulos
UniversityofCalifornia,LosAngeles,CA,USA
DougTygar
UniversityofCalifornia,Berkeley,CA,USA
MosheY.Vardi
RiceUniversity,Houston,TX,USA
GerhardWeikum
MaxPlanckInstituteforInformatics,Saarbruecken,Germany
Marina L. Gavrilova C.J. Kenneth Tan (Eds.)
Transactions on
Computational
Science VIII
1 3
Editors-in-Chief
MarinaL.Gavrilova
UniversityofCalgary,DepartmentofComputerScience
2500UniversityDriveN.W.,Calgary,AB,T2N1N4,Canada
E-mail:[e-mail]
C.J.KennethTan
ExascalaLtd.
Unit9,97RickmanDrive,BirminghamB152AL,UK
E-mail:[e-mail]
LibraryofCongressControlNumber:2010935450
CRSubjectClassification(1998):F.2,F.1,I.4,J.3,I.5,G.2
ISSN 0302-9743(LectureNotesinComputerScience)
ISSN 1866-4733(TransactiononComputationalScience)
ISBN-10 3-642-16235-5SpringerBerlinHeidelbergNewYork
ISBN-13 978-3-642-16235-0SpringerBerlinHeidelbergNewYork
Thisworkissubjecttocopyright.Allrightsarereserved,whetherthewholeorpartofthematerialis
concerned,specificallytherightsoftranslation,reprinting,re-useofillustrations,recitation,broadcasting,
reproductiononmicrofilmsorinanyotherway,andstorageindatabanks.Duplicationofthispublication
orpartsthereofispermittedonlyundertheprovisionsoftheGermanCopyrightLawofSeptember9,1965,
initscurrentversion,andpermissionforusemustalwaysbeobtainedfromSpringer.Violationsareliable
toprosecutionundertheGermanCopyrightLaw.
springer.com
©Springer-VerlagBerlinHeidelberg2010
PrintedinGermany
Typesetting:Camera-readybyauthor,dataconversionbyScientificPublishingServices,Chennai,India
Printedonacid-freepaper 06/3180
LNCS Transactions on Computational Science
Computational science, an emerging and increasingly vital field, is now widely
recognized as an integral part of scientific and technical investigations, affecting
researchers and practitioners in areas ranging from aerospace and automotive research
to biochemistry, electronics, geosciences, mathematics, and physics. Computer
systems research and the exploitation of applied research naturally complement each
other. The increased complexity of many challenges in computational science
demands the use of supercomputing, parallel processing, sophisticated algorithms,
and advanced system software and architecture. It is therefore invaluable to have
input by systems research experts in applied computational science research.
Transactions on Computational Science focuses on original high-quality research
in the realm of computational science in parallel and distributed environments, also
encompassing the underlying theoretical foundations and the applications of large-
scale computation. The journal offers practitioners and researchers the opportunity to
share computational techniques and solutions in this area, to identify new issues, and
to shape future directions for research, and it enables industrial users to apply leading-
edge, large-scale, high-performance computational methods.
In addition to addressing various research and application issues, the journal aims
to present material that is validated – crucial to the application and advancement of
the research conducted in academic and industrial settings. In this spirit, the journal
focuses on publications that present results and computational techniques that are
verifiable.
Scope
The scope of the journal includes, but is not limited to, the following computational
methods and applications:
• Aeronautics and Aerospace
• Astrophysics
• Bioinformatics
• Climate and Weather Modeling
• Communication and Data Networks
• Compilers and Operating Systems
• Computer Graphics
• Computational Biology
• Computational Chemistry
• Computational Finance and Econometrics
• Computational Fluid Dynamics
• Computational Geometry
VI LNCS Transactions on Computational Science
• Computational Number Theory
• Computational Physics
• Data Storage and Information Retrieval
• Data Mining and Data Warehousing
• Grid Computing
• Hardware/Software Co-design
• High-Energy Physics
• High-Performance Computing
• Numerical and Scientific Computing
• Parallel and Distributed Computing
• Reconfigurable Hardware
• Scientific Visualization
• Supercomputing
• System-on-Chip Design and Engineering
Editorial
The Transactions on Computational Science journal is part of the Springer series
Lecture Notes in Computer Science, and is devoted to the gamut of computational
science issues, from theoretical aspects to application-dependent studies and the
validation of emerging technologies.
The journal focuses on original high-quality research in the realm of computational
science in parallel and distributed environments, encompassing the facilitating
theoretical foundations and the applications of large-scale computations and massive
data processing. Practitioners and researchers share computational techniques and
solutions in the area, identify new issues, and shape future directions for research, as
well as enable industrial users to apply the techniques presented.
The current issue is comprised of two parts. Part 1 focuses on adaptive
evolutionary computation, and was prepared by Guest Editors Nadia Nedjah,
Abdelhamid Bouchachia and Luiza de Macedo Mourelle. The focus of Part 2 is on
computational methods for model visualization and analysis.
Part I consists of five manuscripts, each addressing a specific computational
problem utilizing adaptive methodologies. It also contains a comprehensive review of
the current advancements in the area prepared by the Special Issue Guest Editors.
Part II continues the theme. It is comprised of six manuscripts that take an in-depth
look at selected computational science research in the areas of geometric computing,
Euclidean distance transform, distributed systems, segmentation, visualization of
monotone data and data interpolation. Each paper provides a rigorous analysis or a
detailed experiment to amplify the impact of the contribution.
In conclusion, we would like to extend our sincere appreciation to Special Issue
Guest Editors Nadia Nedjah, Abdelhamid Bouchachia and Luiza de Macedo Mourelle
for their efficiency and diligence, all authors for submitting their papers, and all
Associate Editors and referees for their valuable work. We would also like to express
our gratitude to the LNCS editorial staff of Springer, in particular Alfred Hofmann,
Ursula Barth and Anna Kramer, who supported us at every stage of the project.
It is our hope that the fine collection of papers presented in this journal issue will
be a valuable resource for Transactions on Computational Science readers and will
stimulate further research into the vibrant area of computational science applications.
June 2010 Marina L. Gavrilova
C.J. Kenneth Tan
Adaptive Evolutionary Computation
Special Issue Guest Editors’ Preface
Adaptation plays a central role in dynamically changing systems. It is about
the ability ofa systemto “responsively”self-adjustin responseto change inthe
surrounding environment. Like living creatures that have evolved over millions
ofyearsdevelopingecologicalsystemsduetotheirself-adaptationandfitnessca-
pacitytothedynamicenvironment,systemsundergosimilarcyclestoimproveor
atleastto notweakentheir performancewheninternalorexternalchangestake
place.Internalchangebearsonthephysicalstructureofthesystem(thebuilding
blocks:hardwareand/orsoftwarecomponents).Externalchangeoriginatesfrom
the environment due to reciprocal action and interaction. These two classes of
change shed light on the researchavenues towards smart adaptive systems.
Thestateoftheartdrawsthepictureofchallengesthatsuchsystemsneedto
face before they become reality. A sustainable effort is necessary to develop in-
telligent hardware on one level and concepts and algorithms on the other level.
The former level concerns various analog and digital accommodations encom-
passing self-healing, self-testing, reconfigurationand many other aspects of sys-
temdevelopmentandmaintenance.Thelatterlevelisconcernedwithdeveloping
algorithms, concepts and techniques that can rely on metaphors of nature and
that are inspired from biological and cognitive plausibility.
Toface the differenttypes ofchange,systems mustself-adapttheirstructure
and self-adjust their controlling parameters over time as changes are sensed. A
fundamental issue is the notion of “self”, which refers to the capability of the
systemstoactandreactontheirown.Itcoversallstagesofthesystemsworking
and maintenance cycle starting from online self-monitoring to self-growing and
self-organizing.Relying on the two-fold plausibility, which is the basis for many
computationalmodels,neuralnetworkscanbeencounteredinvariousreal-world
dynamicalandnon-stationarysystemsthatrequirecontinuousupdateovertime.
Thereexistmanyneuralmodelsthataretheoreticallybasedonincremental(i.e.,
online, sequential) learning, addressing in particular the notions of self-growing
and self-organizing. However, their strength in practical situations that involve
online adaptation is not as efficient as desirable.
Part I of this journal aims to present the latest advances of adaptive models
for evolutionary computation and their application in various dynamic environ-
ments. The special issue is intended for a wide audience including scientists
as well as mathematicians, physicists, engineers, computer scientists, biologists,
economists and social scientists. This special section covers various topics of
evolutionary computation related to self-organization, self-monitoring and self-
growing concepts. It also aims at presenting a coherentview of these issues and
athoroughdiscussionaboutthe future researchavenues.Themaincontribution
of each of the five papers included is briefly introduced in the following.
X Guest Editor’s Preface
Inthefirstpaperofthesection,entitled“EnvironmentalModelingandIden-
tification Based on Changes in Sensory Information”, the authors present an
environmentalmodelingmethodbasedonstaterepresentation,whichrepresents
a change in sensory information. The model presentedenables the mobile robot
to identify which environment it is in. The results of experiments on a real mo-
bilerobotwithonlylow-sensitivityinfraredsensorsshowthe effectivenessofthe
method.
In the second paper, entitled “Polymorphic Particle Swarm Optimization”,
the authorproposesa newconceptcalledpolymorphicparticleswarmoptimiza-
tion, which generalizes the standard update rule by a polymorphic update rule.
Thispolymorphicupdateruleisanadaptiveupdaterulechangingsymbolsbased
on accumulative histogramsand roulette-wheel sampling. The proposed variant
is applied to typical benchmark functions and in most cases it outperforms the
other PSO existing variants.
In the third paper, entitled “C-Strategy: A Dynamic Adaptive Strategy for
theCLONALGAlgorithm”,the authorsproposeanewparametercontrolstrat-
egy for the immune algorithm CLONALG. The approach presented is based on
the concepts behind reinforcement learning. The approach provides an efficient
andlow costadaptivetechnique forparametercontrol.The results obtainedare
very encouraging.
Inthe fourthpaper,entitled“AComparisonofGenotypeRepresentationsto
AcquireStockTradingStrategyUsingGeneticAlgorithms”,theauthorscompare
somegenotypecodingmethodsoftechnicalindicatorsandtheirparameterstoac-
quireastocktradingstrategyusinggeneticalgorithms.Theyshowthatthesecon-
ventionalrepresentationsarenotsoeffectivefortheGAsearch.Thereafter,they
proposeanewgenotypecodingmethods,namelythe allele-basedindirectrepre-
sentationandshowthesuperiorityoftheproposedindividualrepresentation.
In the fifth paper,entitled “Automatic Adaptive Modeling ofFuzzy Systems
Using Particle Swarm Optimization”, the authors show how to yield adaptive
fuzzy models by applying a particle swarm-based optimization method. The
resultsofthisautomaticmodelingaregivenforthreecomplexthree-dimensional
functions and prove the effectiveness of the proposed method.
The Guest Editors are very grateful to the authors of this special section
and to the reviewers for their tremendous service in critically reviewing the
submitted papers. The editors would also like to thank Prof. Marina Gavrilova,
the Editor-in-Chief of Transactions on Computational Science, Springer-Verlag,
for the editorial assistance and excellent cooperative collaborationin producing
this scientific work. We hope that the reader will share our excitement about
the papers on adaptive evolutionary computation and will find them useful.
June 2010 Nadia Nedjah
Abdelhamid Bouchachia
Luiza de Macedo Mourelle
LNCS Transactions on
Computational Science –
Editorial Board
Marina L. Gavrilova, Editor-in-chief University of Calgary, Canada

Chih Jeng Kenneth Tan, Editor-in-chief OptimaNumerics, UK

Tetsuo Asano JAIST, Japan

Brian A. Barsky University of California at Berkeley, USA

Alexander V. Bogdanov Institute for High Performance Computing

and Data Bases, Russia
Martin Buecker Aachen University, Germany

Rajkumar Buyya University of Melbourne, Australia

Hyungseong Choo Sungkyunkwan University, Korea

Danny Crookes Queen's University Belfast, UK

Tamal Dey Ohio State University, USA

Ivan Dimov Bulgarian Academy of Sciences, Bulgaria

Magdy El-Tawil Cairo University, Egypt
Osvaldo Gervasi Università degli Studi di Perugia, Italy

Christopher Gold University of Glamorgan, UK

Rodolfo Haber Council for Scientific Research, Spain

Andres Iglesias University of Cantabria, Spain

Deok-Soo Kim Hanyang University, Korea

Ivana Kolingerova University of West Bohemia, Czech Republic

Vipin Kumar Army High Performance Computing Research Center, USA

Antonio Lagana Università degli Studi di Perugia, Italy

D.T. Lee Institute of Information Science, Academia Sinica, Taiwan

Laurence Liew Platform Computing, Singapore

Nikolai Medvedev Novosibirsk Russian Academy of Sciences, Russia

Graham M Megson University of Reading, UK

Edward D. Moreno UEA – University of Amazonas state, Brazil

Youngsong Mun Soongsil University, Korea

Dimitri Plemenos Université de Limoges, France

Viktor K. Prasanna University of Southern California, USA

Muhammad Sarfraz KFUPM, Saudi Arabia

Dale Shires Army Research Lab, USA

Masha Sosonkina Ames Laboratory, USA

Alexei Sourin Nanyang Technological University, Singapore

David Taniar Monash University, Australia

Athanasios Vasilakos University of Western Macedonia, Greece

Chee Yap New York University, USA

Igor Zacharov SGI Europe, Switzerland

Zahari Zlatev National Environmental Research Institute, Denmark