Challenges at the interface of data analysis, computer science, and optimization : proceedings of the 34th Annual Conference of the Gesellschaft für Klassifikation e. V., Karlsruhe, July 21-23, 2010

613 Pages·2012·8.414 MB·English

by  Wolfgang A Gaul, Gesellschaft für Klassifikation. Jahrestagung, et al (eds.)

Studies in Classification, Data Analysis,
and Knowledge Organization
ManagingEditors EditorialBoard
(cid:2)
H.-H.Bock,Aachen Ph.Arabie,Newark
W.Gaul,Karlsruhe D.Baier,Cottbus
M.Vichi,Rome F.Critchley,MiltonKeynes
R.Decker,Bielefeld
E.Diday,Paris
M.Greenacre,Barcelona
C.N.Lauro,Naples
J.Meulman,Leiden
P.Monari,Bologna
S.Nishisato,Toronto
N.Ohsumi,Tokyo
O.Opitz,Augsburg
G.Ritter,Passau
M.Schader,Mannheim
C.Weihs,Dortmund
Forfurthervolumes:
http://www.springer.com/series/1564
•
Wolfgang Gaul Andreas Geyer-Schulz
(cid:2)
Lars Schmidt-Thieme Jonas Kunze
(cid:2)
Editors
Challenges at the Interface
of Data Analysis, Computer
Science, and Optimization
Proceedings of the 34th Annual Conference
of the Gesellschaft fu¨r Klassifikation e. V.,
Karlsruhe, July 21 - 23, 2010
123
Editors
Prof.Dr.WolfgangGaul Prof.AndreasGeyer-Schulz
KarlsruheInstituteofTechnology(KIT) KarlsruheInstituteofTechnology(KIT)
InstituteofDecisionTheory InsituteforInformationSystems
andOperationsResearch andManagement(IISM)
Kaiserstr.12 Kaiserstr.12
76128Karlsruhe 76131KarlsruheBaden-Wu¨rttemberg
Germany Germany
[email protected] [email protected]
Prof.Dr.Dr.LarsSchmidt-Thieme JonasKunze
UniversityofHildesheim KarlsruheInstituteofTechnology(KIT)
InstituteofComputerScience InstituteforInformationSystems
MarienburgerPlatz22 andManagement(IISM)
31141Hildesheim Kaiserstraße12
Hildesheim 76128Karlsruhe
Germany Germany
[email protected] [email protected]
ISSN1431-8814
ISBN978-3-642-24465-0 e-ISBN978-3-642-24466-7
DOI10.1007/978-3-642-24466-7
SpringerHeidelbergDordrechtLondonNewYork
LibraryofCongressControlNumber:2012930643
(cid:2)c Springer-VerlagBerlinHeidelberg2012
Thisworkissubjecttocopyright.Allrightsarereserved,whetherthewholeorpartofthematerialis
concerned,specificallytherightsoftranslation,reprinting,reuseofillustrations,recitation,broadcasting,
reproductiononmicrofilmorinanyotherway,andstorageindatabanks.Duplicationofthispublication
orpartsthereofispermittedonlyundertheprovisionsoftheGermanCopyrightLawofSeptember9,
1965,initscurrentversion,andpermissionforusemustalwaysbeobtainedfromSpringer.Violations
areliabletoprosecutionundertheGermanCopyrightLaw.
Theuseofgeneral descriptive names,registered names, trademarks, etc. inthis publication doesnot
imply,evenintheabsenceofaspecificstatement,thatsuchnamesareexemptfromtherelevantprotective
lawsandregulationsandthereforefreeforgeneraluse.
Printedonacid-freepaper
SpringerispartofSpringerScience+BusinessMedia(www.springer.com)
Preface
Revisedversionsofselectedpaperspresentedatthe34thAnnualConferenceofthe
GermanClassificationSociety(GfKl(Gesellschaftfu¨rKlassifikation),amemberof
IFCS(InternationalFederationofClassificationSocieties)),heldatKIT(Karlsruhe
Institute of Technology)in July 2010, are contained in this volume of “Studies in
Classification,DataAnalysis,andKnowledgeOrganization”.
One aimoftheconferencewastoprovidea platformfordiscussionsonresults
concerning the interface that data analysis has in common with other areas such
as, e.g., computer science, operations research, and statistics from a scientific
perspective, as well as with various application areas when “best” interpretations
ofdatathatdescribeunderlyingproblemsituationsneedknowledgefromdifferent
researchdirections.
Practitionersandresearchers–interestedindataanalysisinthebroadsense–had
the opportunityto discuss recent developmentsand to establish cross-disciplinary
cooperationin their fields of interest. More than 200 persons attended the confer-
ence, 158 talks (including plenary and semiplenary lectures) were presented. The
audienceoftheconferencewasquiteinternationalwithabout90contributionsfrom
participants from abroad with the largest groups of foreign presenters from Italy,
Japan, and Poland. Additionally, a program for librarians (14 presentations) was
organized. Parallel to the conference a German-Japanese workshop sponsored by
GfKlandJCS(JapaneseClassificationSociety)tookplacewithadditional24talks.
Sixty of the papers presented at the conference are contained in this volume
(The contributions given at the German-Japanese workshop will be published
elsewhere.). As an unambiguous assignment of topics addressed in single papers
issometimesdifficultthecontributionsaregroupedinawaythattheeditorsfound
appropriate.Within (sub)chaptersthe presentationsare listed in alphabeticalorder
withrespecttotheauthors’names.Attheendofthisvolumeanindexisincluded
that,additionally,shouldhelptheinterestedreader.
Last but not least, we would like to thank all participants of the conference
for their interest and various activities which, again, made the 34th annual GfKl
conferenceandthisvolumeaninterdisciplinarypossibilityforscientificdiscussion,
in particular all authors and all colleagues who reviewed papers, chaired sessions
v
vi Preface
orwereotherwiseinvolved.Here,theteamsofProf.Gaul,Prof.Geyer-Schulz,and
Prof.Schmidt-Thiemehaveprovidedsupportinawaythatoutsidepersonscannot
fully appreciate. Additionally, we gratefully take the opportunity to acknowledge
support by Deutsche Forschungsgemeinschaft (DFG) as well as the Fakulta¨t fu¨r
WirtschaftswissenschaftenofKIT.
This volume was puttogether at Lehrstuhl“InformationServices& Electronic
Markets”, and, at least, Jonas Kunze should be explicitly mentioned for the
final editing tasks. As always we thank Springer Verlag, Heidelberg, especially
Dr.MartinaBihn,forexcellentcooperationinpublishingthisvolume.
KarlsruheandHildesheim(Germany) WolfgangGaul
AndreasGeyer-Schulz
LarsSchmidt-Thieme
JonasKunze
Conference Organization
LocalOrganizers
Prof.Dr.WolfgangGaul(Chair)
Prof.Dr.AndreasGeyer-Schulz
Prof.Dr.MartinE.Ruckes
Prof.Dr.DetlefSeese
Prof.Dr.Karl-HeinzWaldmann
Fakulta¨tfu¨rWirtschaftswissenschaften
KollegiumamSchloss,Universita¨tKarlsruhe(TH),KIT
ScientificProgramCommittee
Y.Baba(Tokyo,Japan)
D.Baier(Cottbus)
H.-H.Bock(Aachen)
A.-L.Boulesteix(Mu¨nchen)
M.P.Brito(Porto,Portugal)
J.M.Buhmann(Zurich,Switzerland)
A.Cerioli(Parma,Italy)
R.Decker(Bielefeld)
L.deRaedt(Leuven,Belgium)
W.Esswein(Dresden)
W.Gaul(Karlsruhe)
M.Greenacre(Barcelona,Spain)
P.J.F.Groenen(Rotterdam,TheNetherlands)
M.Gro¨tschel(Berlin)
Ch.Hennig(London,GreatBritain)
K.Hornik(Vienna,Austria)
O.Hudry(Paris,France)
R.Klein(Augsburg)
vii
viii ConferenceOrganization
P.Kuntz(Nantes,France)
G.McLachlan(Brisbane,Australia)
M.Mizuta(Sapporo,Japan)
A.Okada(Tokyo,Japan)
S.T.Rachev(Karlsruhe)
M.Schader(Mannheim)
L.Schmidt-Thieme(Hildesheim,Chair)
W.Seidel(Hamburg)
M.Spiliopoulou(Magdeburg)
M.Vichi(Rome,Italy)
C.Weihs(Dortmund)
Contents
PartI Classification,ClusterAnalysis,andMultidimensional
Scaling
FuzzificationofAgglomerativeHierarchical
CrispClusteringAlgorithms.................................................... 3
MathiasBankandFriedhelmSchwenker
AnEMAlgorithmfortheStudent-t Cluster-WeightedModeling.......... 13
SalvatoreIngrassia,SimonaC.Minotti,andGiuseppeIncarbone
AnalysisofDistributionValuedDissimilarityData .......................... 23
MasahiroMizutaandHiroyukiMinami
AnOverallIndexforComparingHierarchicalClusterings................. 29
I.MorliniandS.Zani
AnAcceleratedK-MeansAlgorithmBasedonAdaptiveDistances........ 37
Hans-JoachimMuchaandHans-GeorgBartel
Bias-VarianceAnalysisofLocalClassificationMethods .................... 49
JuliaSchiffner,BerndBischl,andClausWeihs
EffectofDataStandardizationontheResultofk-MeansClustering...... 59
KensukeTaniokaandHiroshiYadohisa
ACaseStudyontheUseofStatisticalClassificationMethods
inParticlePhysics................................................................ 69
Claus Weihs, Olaf Mersmann,Bernd Bischl, Arno Fritsch,
HeikeTrautmann,TillMoritzKarbach,and BernhardSpaan
ProblemsofFuzzyc-MeansClusteringandSimilarAlgorithms
withHighDimensionalDataSets .............................................. 79
RolandWinkler,FrankKlawonn,andRudolfKruse
ix