Computational Methods of Feature Selection

Note: Supplemental materials are not guaranteed with Rental or Used book purchases.

ISBN: 9781584888789 | 1584888784
Cover: Hardcover
Copyright: 10/29/2007

Rent

(Recommended)

$92.47

Term

Due

Price

Semester

Dec 13

$97.34

Quarter

Aug 16

$92.47

Short Term

Jul 15

$87.60

*This item is part of an exclusive publisher rental program and requires an additional convenience fee. This fee will be reflected in the shopping cart.
Buy New

Usually Ships in 3-5 Business Days

$132.99
eBook

Available Instantly

Online: 180 Days

Downloadable: 180 Days

$40.20

Due to increasing demands for dimensionality reduction, research on feature selection has deeply and widely expanded into many fields, including computational statistics, pattern recognition, machine learning, data mining, and knowledge discovery. Highlighting current research issues, Computational Methods of Feature Selectionintroduces the basic concepts and principles, state-of-the-art algorithms, and novel applications of this tool. The book begins by exploring unsupervised, randomized, and causal feature selection. It then reports on some recent results of empowering feature selection, including active feature selection, decision-border estimate, the use of ensembles with independent probes, and incremental feature selection. This is followed by discussions of weighting and local methods, such as the ReliefF family, k-means clustering, local feature relevance, and a new interpretation of Relief. The book subsequently covers text classification, a new feature selectionscore, and both constraint-guided and aggressive feature selection. The final section examines applications of feature selection in bioinformatics, including feature construction as well as redundancy-, ensemble-, and penalty-based feature selection. Through a clear, concise, and coherent presentation of topics, this volume systematically covers the key concepts, underlying principles, and inventive applications of feature selection, illustrating how this powerful tool can efficiently harness massive, high-dimensional data and turn it into valuable, reliable information.

Introduction and Background	p. 1
Less Is More	p. 3
Background and Basics	p. 4
Supervised, Unsupervised, and Semi-Supervised Feature Selection	p. 7
Key Contributions and Organization of the Book	p. 10
Part I - Introduction and Background	p. 10
Part II - Extending Feature Selection	p. 11
Part III - Weighting and Local Methods	p. 12
Part IV - Text Classification and Clustering	p. 13
Part V - Feature Selection in Bioinformatics	p. 14
Looking Ahead	p. 15
Unsupervised Feature Selection	p. 19
Introduction	p. 19
Clustering	p. 21
The K-Means Algorithm	p. 21
Finite Mixture Clustering	p. 22
Feature Selection	p. 23
Feature Search	p. 23
Feature Evaluation	p. 24
Feature Selection for Unlabeled Data	p. 25
Filter Methods	p. 26
Wrapper Methods	p. 27
Local Approaches	p. 32
Subspace Clustering	p. 32
Co-Clustering/Bi-Clustering	p. 33
Summary	p. 34
Randomized Feature Selection	p. 41
Introduction	p. 41
Types of Randomizations	p. 42
Randomized Complexity Classes	p. 43
Applying Randomization to Feature Selection	p. 45
The Role of Heuristics	p. 46
Examples of Randomized Selection Algorithms	p. 47
A Simple Las Vegas Approach	p. 47
Two Simple Monte Carlo Approaches	p. 49
Random Mutation Hill Climbing	p. 51
Simulated Annealing	p. 52
Genetic Algorithms	p. 54
Randomized Variable Elimination	p. 56
Issues in Randomization	p. 58
Pseudorandom Number Generators	p. 58
Sampling from Specialized Data Structures	p. 59
Summary	p. 59
Causal Feature Selection	p. 63
Introduction	p. 63
Classical "Non-Causal" Feature Selection	p. 65
The Concept of Causality	p. 68
Probabilistic Causality	p. 69
Causal Bayesian Networks	p. 70
Feature Relevance in Bayesian Networks	p. 71
Markov Blanket	p. 72
Characterizing Features Selected via Classical Methods	p. 73
Causal Discovery Algorithms	p. 77
A Prototypical Causal Discovery Algorithm	p. 78
Markov Blanket Induction Algorithms	p. 79
Examples of Applications	p. 80
Summary, Conclusions, and Open Problems	p. 82
Extending Feature Selection	p. 87
Active Learning of Feature Relevance	p. 89
Introduction	p. 89
Active Sampling for Feature Relevance Estimation	p. 92
Derivation of the Sampling Benefit Function	p. 93
Implementation of the Active Sampling Algorithm	p. 95
Data Generation Model: Class-Conditional Mixture of Product Distributions	p. 95
Calculation of Feature Relevances	p. 96
Calculation of Conditional Probabilities	p. 97
Parameter Estimation	p. 97
Experiments	p. 99
Synthetic Data	p. 99
UCI Datasets	p. 100
Computational Complexity Issues	p. 102
Conclusions and Future Work	p. 102
A Study of Feature Extraction Techniques Based on Decision Border Estimate	p. 109
Introduction	p. 109
Background on Statistical Pattern Classification	p. 111
Feature Extraction Based on Decision Boundary	p. 112
MLP-Based Decision Boundary Feature Extraction	p. 113
SVM Decision Boundary Analysis	p. 114
Generalities About Labeled Vector Quantizers	p. 115
Feature Extraction Based on Vector Quantizers	p. 116
Weighting of Normal Vectors	p. 119
Experiments	p. 122
Experiment with Synthetic Data	p. 122
Experiment with Real Data	p. 124
Conclusions	p. 127
Ensemble-Based Variable Selection Using Independent Probes	p. 131
Introduction	p. 131
Tree Ensemble Methods in Feature Ranking	p. 132
The Algorithm: Ensemble-Based Ranking Against Independent Probes	p. 134
Experiments	p. 137
Benchmark Methods	p. 138
Data and Experiments	p. 139
Discussion	p. 143
Efficient Incremental-Ranked Feature Selection in Massive Data	p. 147
Introduction	p. 147
Related Work	p. 148
Preliminary Concepts	p. 150
Relevance	p. 150
Redundancy	p. 151
Incremental Performance over Ranking	p. 152
Incremental Ranked Usefulness	p. 153
Algorithm	p. 155
Experimental Results	p. 156
Conclusions	p. 164
Weighting and Local Methods	p. 167
Non-Myopic Feature Quality Evaluation with (R)ReliefF	p. 169
Introduction	p. 169
From Impurity to Relief	p. 170
Impurity Measures in Classification	p. 171
Relief for Classification	p. 172
ReliefF for Classification and RReliefF for Regression	p. 175
Extensions	p. 178
ReliefF for Inductive Logic Programming	p. 178
Cost-Sensitive ReliefF	p. 180
Evaluation of Ordered Features at Value Level	p. 181
Interpretation	p. 182
Difference of Probabilities	p. 182
Portion of the Explained Concept	p. 183
Implementation Issues	p. 184
Time Complexity	p. 184
Active Sampling	p. 184
Parallelization	p. 185
Applications	p. 185
Feature Subset Selection	p. 185
Feature Ranking	p. 186
Feature Weighing	p. 186
Building Tree-Based Models	p. 187
Feature Discretization	p. 187
Association Rules and Genetic Algorithms	p. 187
Constructive Induction	p. 188
Conclusion	p. 188
Weighting Method for Feature Selection in K-Means	p. 193
Introduction	p. 193
Feature Weighting in k-Means	p. 194
W-k-Means Clustering Algorithm	p. 197
Feature Selection	p. 198
Subspace Clustering with k-Means	p. 200
Text Clustering	p. 201
Text Data and Subspace Clustering	p. 202
Selection of Key Words	p. 203
Related Work	p. 204
Discussions	p. 207
Local Feature Selection for Classification	p. 211
Introduction	p. 211
The Curse of Dimensionality	p. 213
Adaptive Metric Techniques	p. 214
Flexible Metric Nearest Neighbor Classification	p. 215
Discriminant Adaptive Nearest Neighbor Classification	p. 216
Adaptive Metric Nearest Neighbor Algorithm	p. 217
Large Margin Nearest Neighbor Classifiers	p. 222
Support Vector Machines	p. 223
Feature Weighting	p. 224
Large Margin Nearest Neighbor Classification	p. 225
Weighting Features Increases the Margin	p. 227
Experimental Comparisons	p. 228
Conclusions	p. 231
Feature Weighting through Local Learning	p. 233
Introduction	p. 233
Mathematical Interpretation of Relief	p. 235
Iterative Relief Algorithm	p. 236
Algorithm	p. 236
Convergence Analysis	p. 238
Extension to Multiclass Problems	p. 240
Online Learning	p. 240
Computational Complexity	p. 242
Experiments	p. 242
Experimental Setup	p. 242
Experiments on UCI Datasets	p. 244
Choice of Kernel Width	p. 248
Online Learning	p. 248
Experiments on Microarray Data	p. 249
Conclusion	p. 251
Text Classification and Clustering	p. 255
Feature Selection for Text Classification	p. 257
Introduction	p. 257
Feature Selection Phyla	p. 259
Characteristic Difficulties of Text Classification Tasks	p. 260
Text Feature Generators	p. 261
Word Merging	p. 261
Word Phrases	p. 262
Character N-grams	p. 263
Multi-Field Records	p. 264
Other Properties	p. 264
Feature Values	p. 265
Feature Filtering for Classification	p. 265
Binary Classification	p. 266
Multi-Class Classification	p. 269
Hierarchical Classification	p. 270
Practical and Scalable Computation	p. 271
A Case Study	p. 272
Conclusion and Future Work	p. 274
A Bayesian Feature Selection Score Based on Naive Bayes Models	p. 277
Introduction	p. 277
Feature Selection Scores	p. 279
Posterior Inclusion Probability (PIP)	p. 280
Posterior Inclusion Probability (PIP) under a Bernoulli distribution	p. 281
Posterior Inclusion Probability (PIPp) under Poisson distributions	p. 283
Information Gain (IG)	p. 284
Bi-Normal Separation (BNS)	p. 285
Chi-Square	p. 285
Odds Ratio	p. 286
Word Frequency	p. 286
Classification Algorithms	p. 286
Experimental Settings and Results	p. 287
Datasets	p. 287
Experimental Results	p. 288
Conclusion	p. 290
Pairwise Constraints-Guided Dimensionality Reduction	p. 295
Introduction	p. 295
Pairwise Constraints-Guided Feature Projection	p. 297
Feature Projection	p. 298
Projection-Based Semi-supervised Clustering	p. 300
Pairwise Constraints-Guided Co-clustering	p. 301
Experimental Studies	p. 302
Experimental Study - I	p. 302
Experimental Study - II	p. 306
Experimental Study - III	p. 309
Conclusion and Future Work	p. 310
Aggressive Feature Selection by Feature Ranking	p. 313
Introduction	p. 313
Feature Selection by Feature Ranking	p. 314
Multivariate Characteristic of Text Classifiers	p. 316
Term Redundancy	p. 316
Proposed Approach to Reducing Term Redundancy	p. 320
Stemming, Stopwords, and Low-DF Terms Elimination	p. 320
Feature Ranking	p. 320
Redundancy Reduction	p. 322
Redundancy Removal Algorithm	p. 325
Term Redundancy Tree	p. 326
Experimental Results	p. 326
Summary	p. 330
Feature Selection in Bioinformatics	p. 335
Feature Selection for Genomic Data Analysis	p. 337
Introduction	p. 337
Microarray Data and Challenges	p. 337
Feature Selection for Microarray Data	p. 338
Redundancy-Based Feature Selection	p. 340
Feature Relevance and Redundancy	p. 340
An Efficient Framework for Redundancy Analysis	p. 343
RBF Algorithm	p. 345
Empirical Study	p. 347
Datasets	p. 347
Experimental Settings	p. 349
Results and Discussion	p. 349
Summary	p. 351
A Feature Generation Algorithm with Applications to Biological Sequence Classification	p. 355
Introduction	p. 355
Splice-Site Prediction	p. 356
The Splice-Site Prediction Problem	p. 356
Current Approaches	p. 357
Our Approach	p. 359
Feature Generation Algorithm	p. 359
Feature Type Analysis	p. 360
Feature Selection	p. 362
Feature Generation Algorithm (EGA)	p. 364
Experiments and Discussion	p. 366
Data Description	p. 366
Feature Generation	p. 367
Prediction Results for Individual Feature Types	p. 369
Splice-Site Prediction with FGA Features	p. 370
Conclusions	p. 372
An Ensemble Method for Identifying Robust Features for Biomarker Discovery	p. 377
Introduction	p. 377
Biomarker Discovery from Proteome Profiles	p. 378
Challenges of Biomarker Identification	p. 380
Ensemble Method for Feature Selection	p. 381
Feature Selection Ensemble	p. 383
Results and Discussion	p. 384
Conclusion	p. 389
Model Building and Feature Selection with Genomic Data	p. 393
Introduction	p. 393
Ridge Regression, Lasso, and Bridge	p. 394
Drawbacks of the Lasso	p. 396
The Elastic Net	p. 397
Definition	p. 397
A Stylized Example	p. 399
Computation and Tuning	p. 400
Analyzing the Cardiomypathy Data	p. 402
The Elastic-Net Penalized SVM	p. 404
Support Vector Machines	p. 404
A New SVM Classifier	p. 405
Sparse Eigen-Genes	p. 407
PCA and Eigen-Genes	p. 408
Sparse Principal Component Analysis	p. 408
Summary	p. 409
Index	p. 413
Table of Contents provided by Ingram. All Rights Reserved.

Please wait while the item is added to your bag...

FREE SHIPPING

Amazon no longer offers textbook rentals. We do!

Computational Methods of Feature Selection

Summary

Table of Contents

Supplemental Materials