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

Cover......Page 1
Studies in Classification, Data Analysis,\rand Knowledge Organization......Page 2
Challenges at the Interface of Data Analysis, Computer Science, and Optimization\r......Page 4
ISBN 9783642244650......Page 5
Preface......Page 6
Conference Organization......Page 8
Contents......Page 10
Part I:\rClassification, Cluster Analysis, and\rMultidimensional Scaling......Page 16
1 Introduction......Page 18
2.1 Overall Algorithm......Page 19
2.1.1 Basic Concept......Page 20
2.1.3 Fuzzy Membership......Page 21
2.2.1 Special Subgraph Property......Page 22
2.3 Topic Groups......Page 23
3.2 Methodology......Page 24
4 Conclusion......Page 25
References......Page 26
1 Introduction......Page 28
2 The Cluster-Weighted Modeling......Page 29
3 CWM Parameter Estimation Via the EM Algorithm......Page 32
4 Concluding Remarks......Page 35
References......Page 36
1 Introduction......Page 38
2 Distribution Valued Data and Multidimensional Scaling......Page 39
3 MDS for Distribution Valued Data......Page 40
5 Concluding Remarks......Page 41
References......Page 43
1 Introduction......Page 44
2 The Index and Its Properties......Page 45
3 The Comparison of Two Partitions in k Groups......Page 46
4 The Complement of the Index......Page 48
5 Extension to More than Two Clusterings......Page 49
6 Concluding Remarks......Page 50
References......Page 51
1 Introduction......Page 52
2 K-Means Clustering and (Weighted) Squared Euclidean Distance......Page 54
3 An Accelerated K-Means Method Based on Adaptive Distances......Page 56
4 Application to Ecology......Page 60
References......Page 61
1 Introduction......Page 64
2 Local Approaches to Classification......Page 65
3 Bias-Variance Decomposition of the Misclassification Rate......Page 66
4 Benchmark Study......Page 67
5 Summary and Outlook......Page 71
References......Page 72
Effect of Data Standardization on the Result of k-Means Clustering......Page 74
2 Fraction Norm......Page 75
3 Methods of Standardization......Page 76
4 Simulation Design......Page 77
5 Results......Page 79
6 Conclusion......Page 80
References......Page 81
1 Introduction......Page 84
3 Preprocessing......Page 86
4.1 Results......Page 87
5 Outlook and Conclusion......Page 91
References......Page 92
1 Introduction......Page 94
2 The Algorithms and the Test Environment......Page 95
3 The Effect of High Dimensions......Page 98
4 How to Exploit the Algorithm Parameters to Increase Their Effectiveness......Page 99
5 Conclusions......Page 100
References......Page 101
Part II:\rQuantification Theory......Page 104
1 Introduction......Page 106
2 Total Information Analysis (TIA) or Comprehensive Dual Scaling......Page 107
4 Reduced Versus Total Space Comparisons......Page 109
4.1 Barley Data......Page 110
4.2 Rorschach Data......Page 111
5 Concluding Remarks......Page 113
References......Page 114
1 Introduction......Page 116
2.1 Binary Data Represented as a Hypercube......Page 118
2.2 Individual Scores as Affine Projections......Page 119
2.3 Properties of an Affine Projection of a Data Cube......Page 120
3.1 Proof of the Filled Horseshoe......Page 122
References......Page 123
2 Forty Years of Work in a Nutshell......Page 124
2.3 Missing Responses......Page 125
2.7 Computational Simplification......Page 126
3 New Phase: Cluster Analysis for TIA......Page 127
3.1 Max-Mini Clustering......Page 128
3.2 Minimum Distance Clustering of Between-Set Distance Matrix......Page 129
References......Page 131
Part III:\rAnalysis of m-Mode n-Way and\rAsymmetric Data......Page 136
1 Introduction......Page 138
2 Extending the Model......Page 140
3 Parameter Estimation......Page 141
4 Evaluation of Model Fit......Page 142
5 Illustrative Application......Page 143
References......Page 145
Reconstructing One-Mode Three-way Asymmetric Data for Multidimensional Scaling......Page 148
1 Introduction......Page 149
2 The Method......Page 150
3 An Application......Page 151
4 Conclusion and Outlook......Page 153
References......Page 155
1 Introduction......Page 158
2 The Method......Page 159
3 The Analysis......Page 160
4 Discussion......Page 163
References......Page 165
1 Introduction......Page 166
2.1 Visualization of the Symmetric Part S......Page 167
2.2 Visualization of Asymmetric Part A......Page 168
3 Numerical Example......Page 172
4 Conclusion......Page 173
References......Page 174
Part IV:\rAnalysis of Visual, Spatial, and Temporal\rData......Page 176
1 Introduction......Page 178
3 Temporal Patterns of Population Change......Page 179
4 Relevance of Temporal Patterns......Page 180
5 Clustering and Classification of Temporal Patterns......Page 181
6 Knowledge Discovery......Page 183
7 Discussion......Page 185
References......Page 186
1 Introduction......Page 188
2 p-Adic Numbers......Page 189
3 p-Adic Encoding of Images......Page 190
4 Epipolar Geometry......Page 192
5 Hensel\'s Lemma and RanSaCp......Page 194
6 Discussion and Outlook......Page 195
7 Conclusion......Page 196
References......Page 197
1 Introduction......Page 198
2 Individualized Error Estimation......Page 199
3 IEE with Various Models......Page 201
4 Time Series Classification Using IEE......Page 203
5 Related Work......Page 204
References......Page 205
1 Introduction......Page 208
2 Study Area and Data......Page 209
3 Methodology and Results......Page 210
3.2 Geographical Analysis Machine......Page 213
3.4 Spatial Scan Statistic......Page 214
4 Discussion of the Results and Conclusions......Page 215
References......Page 216
1 Introduction......Page 218
2 Serial Independence Tests Applied to Residuals......Page 219
3 Simulation Study......Page 222
4 Concluding Remarks......Page 225
References......Page 226
Part V:\rNetwork Data, Graphs, and Social\rRelationships......Page 228
1.1 Basic Problem......Page 230
1.2 Communities Based on Degrees......Page 231
2.1 Characterizations via Degree Distribution......Page 233
2.2 Probability Neighborhoods N......Page 234
3.1 A Greedy Approach......Page 235
4 Outlook......Page 236
References......Page 237
1 Introduction......Page 240
2 Modularity Maximization Strategies......Page 241
3 Agglomerative Hierarchical Algorithms......Page 243
4 Optimization Scheme Discussion......Page 244
5 Summary......Page 246
References......Page 247
1 Introduction......Page 248
2 The Metric Conjoint Analysis Model: Factorial Approximation and External Information......Page 249
3 Network Data Matrix as Relational Constraints on Judges......Page 250
4 Example and Validation Study......Page 252
References......Page 255
Clustering Coefficients of Random Intersection Graphs......Page 258
1 Introduction......Page 259
2 Random Intersection Graphs......Page 260
3 Clustering Coefficient of Active Random Intersection Graphs......Page 262
4 Clustering Coefficient of Passive Random Intersection Graphs......Page 265
References......Page 267
1 Introduction......Page 270
2 Social Network Visualization......Page 271
2.2 Dynamic Social Networks......Page 272
3.1 Basis......Page 273
4 Application......Page 275
References......Page 276
1 Introduction......Page 278
2 Inference by Using Fuzzy Information......Page 279
2.1 Algorithm......Page 281
3.1.1 Experimental Setup......Page 282
3.1.2 Performance on Artificial Data......Page 283
4 Conclusion......Page 284
References......Page 285
GIRAN: A Dynamic Graph Interface to Neighborhoods of Related Articles......Page 286
1 Wikipedia: A Directed Graph of Articles......Page 287
2 Calculation of Neighborhoods of Related Articles......Page 288
3.1 Calculation of the Candidate Set......Page 289
4 GIRAN User Interface to Wikipedia......Page 290
5 Related Work, Evaluation, and Outlook......Page 292
References......Page 293
1 Introduction......Page 296
2 Related Work......Page 297
3.1 RQ1: Folksonomy-Based Social KOS......Page 298
3.2 RQ2: Relevance of Power Tags......Page 301
4 Discussion and Outlook......Page 302
References......Page 303
1 Introduction......Page 306
2 Previous Work and Results......Page 307
3 Proof......Page 309
References......Page 314
1 Introduction......Page 316
2 The Community Membership Life-Cycle Model and Related Work......Page 317
3 Operationalization of Community Roles......Page 318
5 Explorative Analysis and Results......Page 320
References......Page 323
1 Introduction......Page 326
2 Basic Ideas of the Abstract Model......Page 327
3 Exemplary Model......Page 329
4 How to Simulate......Page 330
5 Exemplary Simulation......Page 331
6 Conclusions......Page 332
References......Page 333
1 Introduction......Page 334
2 Bots in the Social Web......Page 335
2.1 Bot Detection......Page 336
3.2 Behavioural and Interactional Indicators......Page 337
3.3 A Finite Mixture Analysis of Bot Indicators......Page 339
References......Page 341
Part VI:\rText Mining......Page 344
1 Introduction......Page 346
2.1 Corpus Pre-processing......Page 347
2.3 Low Rank Kernel Matrix Reflecting Word–Word Similarities......Page 348
4 Results......Page 350
5 Conclusion......Page 352
References......Page 353
1 Introduction......Page 356
2 The Original N-Gram Approach by Cavnar & Trenkle......Page 357
3 The Reduced N-Gram Approach......Page 358
3.4 Short Words......Page 359
4 Data Sources, Preparation, and Classification Results......Page 360
References......Page 363
Part VII:\rDimension Reduction......Page 366
1 Introduction......Page 368
2.1 Conducting LSA......Page 369
2.2 Dimension Reduction......Page 370
2.3 Retrieval in Reduced Space......Page 371
3 HOMALS......Page 372
4 HOMALS in IR......Page 374
References......Page 375
1 Introduction......Page 378
2 Definitions......Page 379
3 Empirical Evaluation of Neighbourhood Stability......Page 380
4 Theoretical Assessment......Page 382
5 Discussion......Page 384
References......Page 385
Part VIII:\rStatistical Musicology......Page 386
1 Introduction......Page 388
2.1 Absolute Amplitude Envelope......Page 389
2.2 Pitchless Periodogram......Page 390
2.3 Mel-Frequency Cepstral Coefficients......Page 391
2.4 LPC Simplified Spectral Envelope......Page 392
3 Classification and Evaluation......Page 393
4 Interpretation of Selected Features......Page 394
References......Page 395
The Recognition of Consonance Is not Impaired by Intonation Deviations: A Revised Theory......Page 398
2 Analysis of Stochastic and Periodic Neuronal Activity......Page 399
3 Integration Capability of a Nerve Cell......Page 400
4 Precision and Ambiguity for Coding Periodic Signals......Page 401
5 Experimental Setting and Results: Evidence of Statistical Distribution of Judgements......Page 403
References......Page 405
1 Introduction......Page 408
2.1 Support Vector Machine......Page 409
3.1 Audio Features......Page 410
3.1.2 Groups of features......Page 412
4 Results......Page 413
5 Conclusions and Outlook......Page 414
References......Page 415
1 Introduction......Page 416
2.1 Method Chain......Page 417
2.2 Evaluation Metrics......Page 418
3.1 Setup......Page 419
3.2 Discussion of Results......Page 420
4 Conclusion and Outlook......Page 423
References......Page 424
1 Introduction......Page 426
2.2 Feature Data Reduction......Page 427
3.2 Processing of Partition Based Classification Results......Page 429
4.1 Design of Experiments......Page 430
4.2 Discussion of Results......Page 431
References......Page 433
1 Introduction......Page 436
2 Software Products......Page 437
3 Overall Ranking......Page 444
4 Conclusion......Page 446
References......Page 447
Part IX:\rData Analysis in Banking and Finance......Page 448
1 Introduction......Page 450
2 Data and Methodology......Page 452
3.2 Unconditional Time-Series Regressions......Page 453
3.3 Conditional Time-Series Regressions......Page 455
3.4 Cross-Sectional Regressions......Page 456
References......Page 457
Classification of Large Imbalanced Credit Client Data with Cluster Based SVM......Page 458
1 Motivation and Background......Page 459
2 Extracting Information from Large Data Sets......Page 460
3 Credit Client Data Set Description......Page 461
5 Symbolic Coding of Data Clusters......Page 462
6 Classification Results......Page 463
7 Conclusions......Page 464
References......Page 465
1 Introduction and Related Work......Page 468
2 Peer Group Analysis......Page 469
2.2 Peer Group Size......Page 470
3 Weather Data......Page 471
4.1 Fault Definition......Page 472
4.2 Global Outlier Removal......Page 473
4.3 Peer Group Analysis Experiments......Page 474
References......Page 476
Part X:\rData Analysis in Health and Environment......Page 478
1 Introduction......Page 480
2.2 From Raw Data to Patterns......Page 481
3.2 Mining Models and Techniques......Page 482
4 Working Results and Evaluation......Page 485
5 Conclusions and Future Work......Page 486
References......Page 487
Cluster Analytic Strategy for Identification of Metagenes Relevant for Prognosis of Node Negative Breast Cancer......Page 490
2 Data......Page 491
3 Methods......Page 492
3.1 Cluster Analysis......Page 493
3.3 Metagene Calculation......Page 494
4 Results......Page 495
5 Discussion......Page 496
References......Page 497
Part XI:\rAnalysis of Marketing, Conjoint, and\rMultigroup Data......Page 500
1 Introduction......Page 502
2 Usage of Image Clustering for Marketing Purposes......Page 503
3 Image Clustering Algorithms......Page 506
5 Conclusions and Outlook......Page 508
References......Page 509
1 Introduction......Page 510
2 The Chin/Keil Approach to Multi-Group Analysis......Page 511
3 A New PLS-Based Approach to Multi-Group Analysis......Page 512
4 A Marketing Example......Page 513
5 Discussion......Page 515
References......Page 516
1 Introduction......Page 518
2 Literature Review and Hypothesis......Page 519
3.1 Measurement......Page 520
4 Results......Page 521
References......Page 524
1 Introduction......Page 526
2 The SRF(1) Model with Endogeneity......Page 527
2.1 MCMC Estimation in the SRF(1) Model......Page 530
3 Application: Sales Response in Pharma Marketing......Page 531
3.1 Exogenous Variables: A SRF(1) Model for Regional Pharma Sales......Page 532
4 Summary......Page 533
References......Page 534
1 Introduction......Page 536
2.2 Conjoint Analysis......Page 537
2.3 Lead User Method......Page 538
3.1 Application Outline......Page 539
3.2 Data Collection......Page 540
3.4 Results......Page 541
4 Conclusions and Outlook......Page 542
References......Page 543
1 Introduction......Page 544
2 Problems with Conjoint Analysis Applications......Page 545
3 Known Approaches to Deal with These Problems......Page 546
5 Comparison of the New Approach eACA to ACA......Page 547
References......Page 550
1 Introduction......Page 552
2.1 Research Design......Page 553
2.2 Missing Completely at Random......Page 554
2.3 Missing Completely at Random in Classes......Page 555
References......Page 558
Part XII:\rData Analysis in Education and Psychology......Page 560
1 Stereotypes and Teacher–Student Interaction......Page 562
1.1 Empirical Findings on Student Stereotypes that Teachers Possess......Page 563
2.1 Participants and Procedure......Page 564
3 Results......Page 565
3.1 Interpretation of the Clusters......Page 567
4 Discussion......Page 569
References......Page 570
1 Introduction......Page 572
2 ECTS Grades......Page 573
3 ECTS Grades Assignment Function......Page 574
4 Reference Distribution......Page 577
5 Quality Measures......Page 578
References......Page 579
Students Reading Motivation: A Multilevel Mixture Factor Analysis......Page 582
1 Introduction......Page 583
3 Application to Reading Motivation......Page 584
4 Conclusion......Page 586
References......Page 587
Part XIII:\rAnalysis of Tourism Data......Page 590
1 Introduction......Page 592
2 Methodology......Page 593
3 Data......Page 595
4 Discussion and Conclusions......Page 596
References......Page 599
1 Introduction......Page 600
2 The Concept of Polarization and Its Measurement......Page 601
3 An Index of Asymmetric Polarization for Tourism......Page 603
5 Discussion and Conclusions......Page 605
References......Page 607
Index......Page 610