Multiple Classifier Systems: 12th International Workshop, MCS 2015, Günzburg, Germany, June 29 - July 1, 2015, Proceedings PDF
Preview Multiple Classifier Systems: 12th International Workshop, MCS 2015, Günzburg, Germany, June 29 - July 1, 2015, Proceedings
Friedhelm Schwenker Fabio Roli Josef Kittler (Eds.) 2 Multiple 3 1 9 S Classifier Systems C N L 12th International Workshop, MCS 2015 Günzburg, Germany, June 29 – July 1, 2015 Proceedings 123 Lecture Notes in Computer Science 9132 Commenced Publication in 1973 Founding and Former Series Editors: Gerhard Goos, Juris Hartmanis, and Jan van Leeuwen Editorial Board David Hutchison Lancaster University, Lancaster, UK Takeo Kanade Carnegie Mellon University, Pittsburgh, PA, USA Josef Kittler University of Surrey, Guildford, UK Jon M. Kleinberg Cornell University, Ithaca, NY, USA Friedemann Mattern ETH Zurich, Zürich, Switzerland John C. Mitchell Stanford University, Stanford, CA, USA Moni Naor Weizmann Institute of Science, Rehovot, Israel C. Pandu Rangan Indian Institute of Technology, Madras, India Bernhard Steffen TU Dortmund University, Dortmund, Germany Demetri Terzopoulos University of California, Los Angeles, CA, USA Doug Tygar University of California, Berkeley, CA, USA Gerhard Weikum Max Planck Institute for Informatics, Saarbrücken, Germany More information about this series at http://www.springer.com/series/7412 Friedhelm Schwenker Fabio Roli (cid:129) Josef Kittler (Eds.) Multiple fi Classi er Systems 12th International Workshop, MCS 2015 ü – G nzburg, Germany, June 29 July 1, 2015 Proceedings 123 Editors Friedhelm Schwenker Josef Kittler Ulm University University of Surrey Ulm Guildford Germany UK FabioRoli University of Cagliari Cagliari Italy ISSN 0302-9743 ISSN 1611-3349 (electronic) Lecture Notesin Computer Science ISBN 978-3-319-20247-1 ISBN978-3-319-20248-8 (eBook) DOI 10.1007/978-3-319-20248-8 LibraryofCongressControlNumber:2015940974 LNCSSublibrary:SL6–ImageProcessing,ComputerVision,PatternRecognition,andGraphics SpringerChamHeidelbergNewYorkDordrechtLondon ©SpringerInternationalPublishingSwitzerland2015 Thisworkissubjecttocopyright.AllrightsarereservedbythePublisher,whetherthewholeorpartofthe material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storageandretrieval,electronicadaptation,computersoftware,orbysimilarordissimilarmethodologynow knownorhereafterdeveloped. Theuseofgeneraldescriptivenames,registerednames,trademarks,servicemarks,etc.inthispublication doesnotimply,evenintheabsenceofaspecificstatement,thatsuchnamesareexemptfromtherelevant protectivelawsandregulationsandthereforefreeforgeneraluse. Thepublisher,theauthorsandtheeditorsaresafetoassumethattheadviceandinformationinthisbookare believedtobetrueandaccurateatthedateofpublication.Neitherthepublishernortheauthorsortheeditors give a warranty, express or implied, with respect to the material contained herein or for any errors or omissionsthatmayhavebeenmade. Printedonacid-freepaper SpringerInternationalPublishingAGSwitzerlandispartofSpringerScience+BusinessMedia (www.springer.com) Preface Thisbookpresentstheproceedingsofthe12thIAPRWorkshoponMultipleClassifier Systems (MCS 2015) held at Reisensburg Castle, the research center of Ulm Univer- sity,Germany,duringJune29–July1,2015.TheseriesofMCSworkshopshasacted asamajorforumforinternationalresearchersandpractitionersfromthecommunityof multiple classifier systems in pattern recognition and machine learning. TheProgramCommitteeofMCS2015selected19papersforthescientificprogram. Two IAPR Invited Sessions given by Dr. George Cybenko, Darmouth College, USA, and Dr. Marcello Pelillo, University of Venice, Italy, enriched the workshop. MCS 2015 would not have been possible without the help and support of many peopleandorganizations.Firstofall,wearegratefultoallauthorswhosubmittedtheir work to MCS. We thank the members of the Program Committee and the additional reviewersforperformingthedifficulttaskofselectingthebestpapersforpresentation, and we hope that readers of this volume will enjoy it and be inspired from its contributions. MCS 2015 was supported by the International Association for Pattern Recognition (IAPR), by the University of Cagliari, Italy, by the University of Surrey, UK, and by UlmUniversity,Germany,whichhostedthisevent.Specialthankstothepeopleofthe localorganization,MiriamSchmidt,MartinSchels,MichaelGlodek,MarkusKächele, SaschaMeudt,andPatrickThiam.Finally,wewishtoexpressourgratitudetoSpringer for publishing our proceedings in their LNCS series and for their constant support. July 2015 Friedhelm Schwenker Fabio Roli Josef Kittler Organization Organizing Committee Friedhelm Schwenker University of Ulm, Germany Fabio Roli University of Cagliari, Italy Josef Kittler University of Surrey, UK Program Committee Jon Benediktsson, Iceland Gonzalo Martinez-Munoz, Spain Gavin Brown, UK Juan Jose Rodriguez, Spain Cesare Furlanello, Italy Arun Ross, USA Giorgio Fumera, Italy Carlo Sansone, Italy Joydeep Ghosh, USA Giorgio Valentini, Italy Larry Hall, USA Terry Windeatt, UK Tin Kam Ho, USA Xin Yao, China Philip Kegelmeyer, USA Yang Yu, China Ludmila Kuncheva, UK Zhi-Hua Zhou, China Sponsoring Institutions University of Ulm, Germany University of Cagliari, Italy University of Surrey, UK International Association for Pattern Recognition (IAPR) Deep Learning of Behaviors George Cybenko Thayer Schoolof Engineering Dartmouth College Hanover NH03755, USA [email protected] Abstract. Deep learning has generated much research and commercialization interest recently. In a way, it is the third incarnation of neural networks as pattern classifiers, using insightful algorithms and architectures that act as unsupervisedauto-encoders, learning hierarchies offeatures in adataset. Afterashortreviewofthatwork,wewilldiscussthechallengesassociated withtheanalysisofbehaviorsobservedastimeseriesofcategoricaldata.Novel computational approaches for deep learning of behaviors as opposed to just static patterns will be presented. Our approach is based on structured non- negativematrixfactorizationsofmatricesthatencodeobservationfrequenciesof behaviors. These techniques can be used to robustly characterize and exploit diverse behaviorsinsecurityapplicationssuchascovertchanneldetectionandcoding. Examples ofsuchapplications will bepresented. Relatedresultsabouttheroleofdiversityincomputersecurityapplications willalsobeintroducedwhereinadversarialdynamicsdictatesthatattackersand defenders coevolve. As a result, the use of multiple diverse detection and mitigation techniques makes the attackers’effective workfactor much higher. Similarity-Based Pattern Recognition: A Game-Theoretic Perspective Marcello Pelillo Ca’Foscari University of Venice 30172Venezia Mestre, Italy [email protected] Abstract.Similarity-based methodsareemergingasapowerfultoolinpattern recognition and machine learning because of their ability to overcome the intrinsiclimitationsoftraditionalfeature-vectorapproaches.Bydepartingfrom vectorspace representations, however, one is confronted with the challenging problem of dealing with (dis)similarities that do not necessarily possess the Euclideanbehaviorornotevenobeytherequirementsofametric.Inthistalk, I will maintain that game theory offers an elegant and powerful conceptual framework which serves well this purpose, and I will describe recent attempts aimedatformulatingvarioussimilarity-basedpatternrecognitionproblemsfrom agame-theoreticperspective.Particularemphasiswillbegiventoevolutionary- based models which, in contrast to the classical theory, offer an intriguing dynamicalsystemperspective.Finally,Iwilldescrivesomeapplicationsofthis approach within the contextof multiple classifier systems. Contents Theory and Algorithms A Novel Bagging Ensemble Approach for Variable Ranking and Selection for Linear Regression Models. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Chun-Xia Zhang, Jiang-She Zhang, and Guan-Wei Wang A Hierarchical Ensemble Method for DAG-Structured Taxonomies. . . . . . . . 15 Peter N. Robinson, Marco Frasca, Sebastian Köhler, Marco Notaro, Matteo Re, and Giorgio Valentini Diversity Measures and Margin Criteria in Multi-class Majority Vote Ensemble. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Ayako Mikami, Mineichi Kudo, and Atsuyoshi Nakamura Fractional Programming Weighted Decoding for Error-Correcting Output Codes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 Firat Ismailoglu, I.G. Sprinkhuizen-Kuyper, Evgueni Smirnov, Sergio Escalera, and Ralf Peeters Instance-Based Decompositions of Error Correcting Output Codes . . . . . . . . 51 Firat Ismailoglu, Evgueni Smirnov, Nikolay Nikolaev, and Ralf Peeters Pruning Bagging Ensembles with Metalearning. . . . . . . . . . . . . . . . . . . . . . 64 Fábio Pinto, Carlos Soares, and João Mendes-Moreira Multi-label Selective Ensemble. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 Nan Li, Yuan Jiang, and Zhi-Hua Zhou Supervised Selective Combination of Diverse Object-Representation Modalities for Regression Estimation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 Olga Krasotkina, Oleg Seredin, and Vadim Mottl Detecting Ordinal Class Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 RaphaelLattke,LudwigLausser,ChristophMüssel,andHansA.Kestler Calibrating AdaBoost for Asymmetric Learning . . . . . . . . . . . . . . . . . . . . . 112 Nikolaos Nikolaou and Gavin Brown Building Classifier Ensembles Using Greedy Graph Edit Distance. . . . . . . . . 125 Kaspar Riesen, Miquel Ferrer, and Andreas Fischer
The list of books you might like
