Advanced Studies in Classification and Data Science

Preface......Page 6
Acknowledgements......Page 7
Contents......Page 8
About the Editors......Page 12
Part I Classification and Cluster Analysis......Page 13
1 Introduction......Page 14
2 Multilevel Latent Class Model......Page 16
3 Proposed Latent Variable Prediction Algorithm......Page 18
4 Simulation Study......Page 20
4.1 Study I......Page 21
4.2 Study II......Page 22
5.1 Example 1: Application to INVALSI Data......Page 23
5.2 Example 2: Application to TIMSS&PIRLS Italian Data......Page 24
6 Conclusions......Page 25
Appendix......Page 26
References......Page 27
Multi-Criteria Classifications in Regional Development Modelling......Page 29
2 The Assumptions and Research Procedure......Page 30
3 Analysis of the Results......Page 34
References......Page 38
1 Introduction: Parametric and Non-parametric Mixtures......Page 40
2 EM Latent Clustering: A Concise Derivation from First Principles......Page 42
2.1 Latent Co-clustering......Page 43
2.2.1 Latent Co-clustering......Page 44
2.2.2 Latent Clustering......Page 45
2.3 Case Study I: Reuters 21578 Term-Document Matrix......Page 46
3 Network Clustering......Page 47
3.2 Case Study II: Inter-Cantonal Swiss Migrations......Page 48
3.4 Case Study III: Modeling Bigrams......Page 50
References......Page 52
1 Introduction......Page 54
2.1 Multiple-Kernel Learning Method......Page 57
2.2 Self-Tuning Kernel Learning Method......Page 58
3 Experiments......Page 61
References......Page 64
Random Forests Followed by Computed ABC Analysis as a Feature Selection Method for Machine Learning in Biomedical Data......Page 66
2.1  Biomedical Data Set......Page 67
2.2 Data Preprocessing......Page 68
2.3 Random Forest and Computed ABC Analysis Based Feature Selection......Page 69
2.4 Comparative Assessment of Classifier Performance......Page 70
3.1 Classification Performance Using the Full and Reduced Feature Sets......Page 71
3.3 Reassessment Using Features Unrelated to the Classes......Page 73
4 Discussion......Page 74
References......Page 76
1 Introduction......Page 79
2 The Method......Page 81
3 The Analysis......Page 82
4 Conclusion......Page 84
References......Page 86
1 Introduction......Page 87
2 The Algorithm......Page 89
3 Simulation Study......Page 92
4 Conclusions......Page 95
References......Page 96
1 Introduction......Page 97
3 Problems with the Measurability of Social Processes......Page 98
4.1 Changes in the RCI Over Time......Page 99
4.2 Composition of the Current RCI......Page 100
5.1 The Characteristics of Clusters......Page 102
5.2 Homogeneity......Page 105
6 Final Remarks......Page 108
References......Page 109
1 Introduction......Page 110
2 Related Literature......Page 111
3 Estimation Methods Based on Weighting Clusters......Page 112
3.2 Methodology......Page 113
3.2.1 Cluster Analysis......Page 114
3.2.2 Ranking Classifications......Page 115
3.2.3 Developing a Weight System......Page 116
4 Result......Page 117
References......Page 119
Part II Multidimensional Analysis and Visualization......Page 121
1 Introduction......Page 122
2.2 Pearson\'s Residuals......Page 124
4.1 A General Standardised Residual......Page 126
4.2 Generalised Poisson Distribution......Page 127
4.3 Negative Binomial Distribution......Page 128
5 Variance Stabilising Strategies......Page 129
5.2 Freeman–Tukey Residual......Page 130
6 Discussion......Page 132
References......Page 133
From Joint Graphical Display to Bi-Modal Clustering: [2] Dual Space Versus Total Space......Page 135
2 Exploratory Clustering in Dual Space and Total Space......Page 136
2.1 Example 1: Rorschach Responses Under Different Moods......Page 137
2.2 Example 2: A Subset of Rorschach Responses Under Different Moods......Page 143
3 Concluding Remarks......Page 145
References......Page 147
1 Introduction......Page 148
2.1 Pixellating the Mapped Data Point Cloud......Page 150
3 Visualization Through 2D Histogram Representation......Page 151
3.1 Varying Number Theoretic Representations of the Data: m-Ary and p-Adic Representations......Page 152
4 Hashing and Binning for Nearest Neighbour Searching......Page 154
6 Conclusion......Page 156
References......Page 157
1 To Begin With......Page 159
3 Early Publications in English......Page 160
4 Principal Coordinates and Standard Coordinates......Page 162
5 Joint Graphical Display as a Problematic Approach......Page 163
6 Total Information Analysis or Comprehensive Dual Scaling......Page 164
7 Contingency Table and Response-Pattern Table......Page 165
7.1 Justification of Total Space......Page 166
References......Page 167
1 Introduction......Page 171
2 Biplots for Mixed Types of Data......Page 172
2.2 Representation of Continuous Variables......Page 173
2.3 Representation of Binary Variables......Page 176
2.4 Representation of Nominal Variables......Page 178
2.5 Representation of Ordinal Variables......Page 179
3 Illustrative Example......Page 181
4 Software Note......Page 183
References......Page 184
Part III Statistical Methods......Page 186
1 Introduction......Page 187
2.1 Functionalization......Page 188
2.3 Measure of Similarity Between Dendrograms and Visualization......Page 190
3 Analysis of Concentration Data......Page 191
References......Page 198
1 Introduction......Page 199
2.1 Pre-processing......Page 201
2.2 A Generalized Additive Model for CNV Detection......Page 203
2.3 Final CNV Calling......Page 204
3 Results......Page 208
3.1 Determination of Sample Outliers......Page 209
4 Discussion......Page 211
References......Page 212
1 Introduction......Page 214
2 Classifiers......Page 215
2.4 Multinomial Logistic Regression......Page 216
3 Functional Data......Page 217
4 Functional Distance Covariance and Distance Correlation......Page 218
5 Variable Selection Based on the Distance Covariance......Page 220
6 Real Example......Page 221
References......Page 223
1 Introduction......Page 225
2 The ALS Algorithm for NMVA......Page 226
3.1 The ALS Algorithm for NPCA......Page 228
3.2 Initial Value Selection Procedure of the ALS Algorithm......Page 229
4 Acceleration of the ALS Algorithm in the Initial Value Selection Procedure......Page 231
5 Numerical Experiments......Page 233
6 Discussion......Page 236
References......Page 237
1 Introduction......Page 238
2 Asymptotic Tests......Page 240
3 Permutation and Bootstrap Tests......Page 242
4 Confidence Regions......Page 244
5.1 Simulation Setup......Page 245
5.2 Discussion of the Simulation Results......Page 246
References......Page 250
1 Introduction......Page 252
2 Gap Adjusted Euclidean Distance in R1......Page 253
3 Euclidean Distance in R1 Adjusted for Gap and Outside Neighbourhood......Page 255
4 Tube Distance......Page 256
5 Closing Remarks......Page 260
References......Page 261
1.1 Background......Page 262
1.2 Statistical Matching Methods......Page 263
1.3 Proposed Method......Page 264
2 Model......Page 265
3.1 Data......Page 266
3.3 Matching Performance Evaluation......Page 268
4 Conclusion......Page 270
References......Page 271
Constructing Graphical Models for Multi-Source Data: Sparse Network and Component Analysis......Page 272
1 Background......Page 273
2 Methods......Page 274
3 Simulation Study......Page 279
4 Results......Page 281
5 Conclusion and Discussion......Page 282
References......Page 284
1 Introduction......Page 285
2 Normalized Mutual Information......Page 287
3 Decompositions......Page 288
4 Numerical Examples......Page 290
5 More Numerical Examples......Page 293
References......Page 294
1 Introduction......Page 296
2 Notation......Page 298
3 Rand Index and a Decomposition......Page 299
4 Numerical Examples......Page 301
5 Relative Weights......Page 303
6 Statistical Independence......Page 306
References......Page 307
Layered Multivariate Regression with Its Applications......Page 309
1 Introduction......Page 310
2.1 Algorithm......Page 312
2.2 Layer Number Selection......Page 314
3 Numerical Simulation......Page 315
5 Extension of LMR: Layered Principal Component Analysis......Page 317
6 Concluding Remarks......Page 319
References......Page 320
Part IV Statistical Data Analysis......Page 321
An Exploratory Study on the Clumpiness Measure of Intertransaction Times: How Is It Useful for Customer Relationship Management?......Page 322
2 RFM Measure......Page 323
3 Clumpiness Measure......Page 325
4.1 Data......Page 326
4.2 Method......Page 327
5 Result......Page 329
6 Conclusion......Page 330
References......Page 331
1 Introduction......Page 333
2 Problem Setting......Page 334
2.3 Research Method......Page 335
3 Outlier Detection Algorithms......Page 336
5 Experiment......Page 337
5.2 Implement......Page 338
5.2.4 Step 4: Calculate Algorithm Accuracy......Page 339
5.3.1 Result about Algorithm......Page 341
6 Discussion......Page 342
7 Conclusion......Page 344
References......Page 345
1 Introduction......Page 346
2 Methodology......Page 347
3 Data......Page 348
4.1 Estimation of Weights by Regression Analysis......Page 350
5.1 Comparison of Income and Expenditure of Virtual Households and Actual Households......Page 351
References......Page 355
1 Introduction......Page 356
2 The Administrative Areas of Vietnam......Page 357
3 Data......Page 358
4.1.1 Regional Patterns of Climate......Page 359
4.1.2 Regional Risk of Natural Disasters......Page 360
4.2 Household Income and Poverty in Vietnam......Page 361
5.2 A Quantile Regression: Effects of Natural Disasters on Household Income and Poverty......Page 363
References......Page 367
1 Introduction......Page 368
3 Data......Page 369
5 Results......Page 370
5.2 Path From Num. of Tweets to Sales Performance......Page 371
5.3 Indirect Effect......Page 372
6 Conclusion......Page 373
References......Page 374
1 Introduction......Page 375
3 Reduction and Representation: Cluster Differences Scaling......Page 377
4.1 Distance-Based Linear Model......Page 379
4.2 Distance-Based Generalised Linear Model......Page 380
5 Actuarial Application......Page 381
5.1 Fitting Distance-Based Regression......Page 382
5.2 Fitting Cluster Distance-Based Regression......Page 383
Supporting Agencies......Page 386
References......Page 387
Bayesian Network Analysis of Fashion Behavior......Page 388
1 Introduction......Page 389
3 Explanatory Analysis of Fashion Behavior......Page 390
3.1 Data......Page 391
3.2 Training the Bayesian Network......Page 392
3.3.1 Model Selection......Page 393
3.3.2 Implication of the Networks......Page 395
4 Conclusion......Page 399
References......Page 400
Part V Statistical Data Analysis for Social Science......Page 401
1 Methodology......Page 402
2 Issue Voting Behavioral Map: The Greek National Elections of 2015......Page 409
Appendix......Page 415
References......Page 416
1 Introduction......Page 417
2 Beyond GDP......Page 418
3 Well-Being Measures......Page 419
4 Evaluation Criteria......Page 425
References......Page 428
1 Introduction......Page 430
2 Data and Descriptive Statistics......Page 431
3 Methodology......Page 433
4 Key Results......Page 435
References......Page 443
1 Introduction......Page 444
2 Data......Page 445
3.1 OLS......Page 447
3.2 Moran\'s I Statistics......Page 448
3.3 CAR......Page 451
References......Page 454
1.1 Background......Page 455
1.2.1 Human Resource Hypothesis......Page 456
1.2.2 Supplement Household Income Hypothesis......Page 458
2.1 Data......Page 459
2.2.1 Dependent Variable......Page 460
2.3 Method......Page 461
3.1 Descriptive Statistics......Page 462
3.2 Determinants of Wife\'s Participation in Paid Work......Page 463
4 Conclusion and Discussion......Page 465
References......Page 466
1 Introduction......Page 468
2 The Background Information and Methodology of the Research......Page 469
3 Empirical Analysis Results......Page 470
4 Conclusions......Page 476
References......Page 477
1 Introduction......Page 478
2 About IPUMS......Page 479
3 Harmonization Overview......Page 480
4 Data Standardization......Page 481
5 Variable Harmonization......Page 482
7 Harmonization Challenges......Page 485
8 Summary......Page 493
References......Page 494
1 Introduction......Page 495
2 Autocoding Systems......Page 496
3 Autocoding System Using Simple Classifier Based on Machine Learning......Page 497
4.1 Setting......Page 499
4.2 Results......Page 500
5 Conclusions......Page 505
References......Page 506