9781441912794

Data Mining and information Systems: Quo Vadis?	p. 1
Introduction	p. 1
Special Issues in Data Mining	p. 3
Confirmatory Data Analysis	p. 3
Knowledge Discovery from Supervised Learning	p. 4
Classification Analysis	p. 6
Hybrid Data Mining Procedures	p. 8
Web Mining	p. 10
Privacy-Preserving Data Mining	p. 11
Conclusion and Outlook	p. 12
References	p. 13
Confirmatory Data Analysis
Response-Based Segmentation Using Finite Mixture Partial Least Squares	p. 19
Introduction	p. 20
On the Use of PLS Path Modeling	p. 20
Problem Statement	p. 22
Objectives and Organization	p. 23
Partial Least Squares Path Modeling	p. 24
Finite Mixture Partial Least Squares Segmentation	p. 26
Foundations	p. 26
Methodology	p. 28
Systematic Application of Fimix-Pls	p. 31
Application of Fimix-Pls	p. 34
On Measuring Customer Satisfaction	p. 34
Data and Measures	p. 34
Data Analysis and Results	p. 36
Summary and Conclusion	p. 44
References	p. 45
Knowledge Discovery from Supervised Learning
Building Acceptable Classification Models
Introduction	p. 54
Comprehensibility of Classification Models	p. 55
Measuring Comprehensibility	p. 57
Obtaining Comprehensible Classification Models	p. 58
Justifiability of Classification Models	p. 59
Taxonomy of Constraints	p. 60
Monotonicity Constraint	p. 62
Measuring Justifiability	p. 63
Obtaining Justifiable Classification Models	p. 68
Conclusion	p. 70
References	p. 71
Mining Interesting Rules Without Support Requirement: A General Universal Existential Upward Closure Property	p. 75
Introduction	p. 76
State of the Art	p. 77
An Algorithmic Property of Confidence	p. 80
On UEUC Framework	p. 80
The UEUC Property	p. 80
An Efficient Pruning Algorithm	p. 81
Generalizing the UEUC Property	p. 82
A Framework for the Study of Measures	p. 84
Adapted Functions of Measure	p. 84
Expression of a Set of Measures of Ddconf	p. 87
Conditions for Gueuc	p. 90
A Sufficient Condition	p. 90
A Necessary Condition	p. 93
Classification of the Measures	p. 92
Conclusion	p. 94
References	p. 95
Classification Techniques and Error Control in Logic Mining	p. 99
Introduction	p. 100
Brief Introduction to Box Clustering	p. 102
BC-Based Classifier	p. 104
Best Choice of a Box System	p. 108
Bi-criterion Procedure for BC-Based Classifier	p. 111
Examples	p. 112
The Data Sets	p. 112
Experimental Results with BC	p. 113
Comparison with Decision Trees	p. 115
Conclusions	p. 117
References	p. 117
Classification Analysis
An Extended Study of the Discriminant Random Forest	p. 123
Introduction	p. 123
Random Forests	p. 124
Discriminant Random Forests	p. 125
Linear Discriminant Analysis	p. 126
The Discriminant Random Forest Methodology	p. 127
DRF and RF: An Empirical Study	p. 128
Hidden Signal Detection	p. 129
Radiation Detection	p. 132
Significance of Empirical Results	p. 136
Small Samples and Early Stopping	p. 137
Expected Cost	p. 143
Conclusions	p. 143
References	p. 145
Prediction with the SVM Using Test Point Margins	p. 147
Introduction	p. 147
Methods	p. 151
Data Set Description	p. 154
Results	p. 154
Discussion and Future Work	p. 155
References	p. 157
Effects of Oversampling Versus Cost-Sensitive Learning for Bayesian and SVM Classifiers	p. 159
Introduction	p. 159
Resampling	p. 161
Random Oversampling	p. 161
Generative Oversampling	p. 161
Cost-Sensitive Learning	p. 162
Related Work	p. 163
A Theoretical Analysis of Oversampling Versus Cost-Sensitive Learning	p. 164
Bayesian Classification	p. 164
Resampling Versus Cost-Sensitive Learning in Bayesian Classifiers	p. 165
Effect of Oversampling on Gaussian Naive Bayes	p. 166
Effects of Oversampling for Multinomial Naive Bayes	p. 168
Empirical Comparison of Resampling and Cost-Sensitive Learning	p. 170
Explaining Empirical Differences Between Resampling and Cost-Sensitive Learning	p. 170
Naive Bayes Comparisons on Low-Dimensional Gaussian'Data	p. 171
Multinomial Naive Bayes	p. 176
SVMs	p. 178
Discussion	p. 181
Conclusion	p. 182
Appendix	p. 183
References	p. 190
The Impact of Small Disjuncts on Classifier Learning	p. 193
Introduction	p. 193
An Example: The Vote Data Set	p. 195
Description of Experiments	p. 197
The Problem with Small Disjuncts	p. 198
The Effect of Pruning on Small Disjuncts	p. 202
The Effect of Training Set Size on Small Disjuncts	p. 210
The Effect of Noise on Small Disjuncts	p. 213
The Effect of Class Imbalance on Small Disjuncts	p. 217
Related Work	p. 220
Conclusion	p. 223
References	p. 225
Hybrid Data Mining Procedures
Predicting Customer Loyalty Labels in a Large Retail Database: A Case Study in Chile	p. 229
Introduction	p. 229
Related Work	p. 231
Objectives of the Study	p. 233
Supervised and Unsupervised Learning	p. 234
Unsupervised Algorithms	p. 234
Variables for Segmentation	p. 238
Exploratory Data Analysis	p. 239
Results of the Segmentation	p. 240
Results of the Classifier	p. 241
Business Validation	p. 244
In-Store Minutes Charges for Prepaid Cell Phones	p. 245
Distribution of Products in the Store	p. 246
Conclusions and Discussion	p. 248
Appendix	p. 250
References	p. 252
PCA-Based Time Series Similarity Search	p. 255
Introduction	p. 256
Background	p. 258
Review of PCA	p. 258
Implications of PCA in Similarity Search	p. 259
Related Work	p. 261
Proposed Approach	p. 263
Experimental Methodology	p. 265
Data Sets	p. 265
Evaluation Methods	p. 266
Rival Measures	p. 267
Results	p. 268
I-NN Classification	p. 268
k-NN Similarity Search	p. 271
Speeding Up the Calculation of APEdist	p. 272
Conclusion	p. 274
References	p. 274
Evolutionary Optimization of Least-Squares Support Vector Machines	p. 277
Introduction	p. 278
Kernel Machines	p. 278
Least-Squares Support Vector Machines	p. 279
Kernel Functions	p. 280
Evolutionary Computation	p. 281
Genetic Algorithms	p. 281
Evolution Strategies	p. 282
Genetic Programming	p. 283
Related Work	p. 283
Hyperparameter Optimization	p. 284
Combined Kernel Functions	p. 284
Evolutionary Optimization of Kernel Machines	p. 286
Hyperparameter Optimization	p. 286
Kernel Construction	p. 287
Objective Function	p. 288
Results	p. 289
Data Sets	p. 289
Results for Hyperparameter Optimization	p. 290
Results for EvoKMGP	p. 293
Conclusions and Future Work	p. 294
References	p. 295
Genetically Evolved kNN Ensembles	p. 299
Introduction	p. 299
Background and Related Work	p. 301
Method	p. 302
Data sets	p. 305
Results	p. 307
Conclusions	p. 312
References	p. 313
Web-Mining
Behaviorally Founded Recommendation Algorithm for Browsing Assistance Systems	p. 317
Introduction	p. 317
Related Works	p. 318
Our Contribution and Approach	p. 319
Concept Formalization	p. 319
System Design	p. 323
A Priori Knowledge of Human-System Interactions	p. 323
Strategic Design Factors	p. 323
Recommendation Algorithm Derivation	p. 325
Practical Evaluation	p. 327
Intranet Portal	p. 328
System Evaluation	p. 330
Practical Implications and Limitations	p. 331
Conclusions and Future Work	p. 332
References	p. 333
Using Web Text Mining to Predict Future Events: A Test of the Wisdom of Crowds Hypothesis	p. Scott Ryan
Introduction	p. 335
Method	p. 337
Hypotheses and Goals	p. 337
General Methodology	p. 339
The 2006 Congressional and Gubernatorial Elections	p. 339
Sporting. Events and Reality Television Programs	p. 340
Movie Box Office Receipts and Music Sales	p. 341
Replication	p. 342
Results and Discussion	p. 343
The 2006 Congressional and Gubernatorial Elections	p. 343
Sporting Events and Reality Television Programs	p. 345
Movie and Music Album Results	p. 347
Conclusion	p. 348
References	p. 349
Privacy-Preserving Data Mining
Avoiding Attribute Disclosure with the (Extended) p-Sensitive k-Anonymity Model	p. 353
Introduction	p. 353
Privacy Models and Algorithms	p. 354
The p-Sensitive k-Anonymity Model and Its Extension	p. 354
Algorithms for the p-Sensitive k-Anonymity Model	p. 357
Experimental Results	p. 360
Experiments for p-Sensitivek-Anonymity	p. 360
Experiments for Extended p-Sensitive k-Anonymity	p. 362
New Enhanced Models Based on p-Sensitive k-Anonymity	p. 366
Constrained p-Sensitive k-Anonymity	p. 366
p-Sensitive k-Anonymity in Social Networks	p. 370
Conclusions and Future Work	p. 372
References	p. 372
Privacy-Preserving Random Kernel Classification of Checkerboard Partitioned Data	p. 375
Introduction	p. 375
Privacy-Preserving Linear Classifier for Checkerboard Partitioned Data	p. 379
Privacy-Preserving Nonlinear Classifier for Checkerboard Partitioned Data	p. 381
Computational Results	p. 382
Conclusion and Outlook	p. 384
References	p. 386
Table of Contents provided by Ingram. All Rights Reserved.

Amazon no longer offers textbook rentals. We do!

Amazon no longer offers textbook rentals. We do!

We're the #1 textbook rental company. Let us show you why.

Data Mining

1441912797

Supplemental Materials

Summary

Table of Contents

Supplemental Materials

Rewards Program