Ebook: Feature Extraction: Foundations and Applications

This book is both a reference for engineers and scientists and a teaching resource, featuring tutorial chapters and research papers on feature extraction.
"This book compiles some very promising techniques, coming from an extremely smart collection of researchers, delivering their best ideas in a competitive environment."
Trevor Hastie, Stanford University
"Feature selection is a key technology for making sense of the high dimensional data. Isabelle Guyon et al. have done a splendid job in designing a challenging competition, and collecting the lessons learned."
Bernhard Schoelkopf, Max Planck Institute
"There has been until now insufficient consideration of feature selection algorithms, no unified presentation of leading methods, and no systematic comparisons. This volume is noteworthy for the breadth of methods covered, the clarity of presentations, the unity in notation and the helpful statistical appendices."
David G. Stork, Ricoh Innovations
"Feature extraction finds application in biotechnology, industrial inspection, the Internet, radar, sonar, and speech recognition. This book will make a difference to the literature on machine learning."
Simon Haykin, Mc Master University
"This book sets a high standard as the public record of an interesting and effective competition."
Peter Norvig, Google Inc.

---

This book is both a reference for engineers and scientists and a teaching resource, featuring tutorial chapters and research papers on feature extraction.
"This book compiles some very promising techniques, coming from an extremely smart collection of researchers, delivering their best ideas in a competitive environment."
Trevor Hastie, Stanford University
"Feature selection is a key technology for making sense of the high dimensional data. Isabelle Guyon et al. have done a splendid job in designing a challenging competition, and collecting the lessons learned."
Bernhard Schoelkopf, Max Planck Institute
"There has been until now insufficient consideration of feature selection algorithms, no unified presentation of leading methods, and no systematic comparisons. This volume is noteworthy for the breadth of methods covered, the clarity of presentations, the unity in notation and the helpful statistical appendices."
David G. Stork, Ricoh Innovations
"Feature extraction finds application in biotechnology, industrial inspection, the Internet, radar, sonar, and speech recognition. This book will make a difference to the literature on machine learning."
Simon Haykin, Mc Master University
"This book sets a high standard as the public record of an interesting and effective competition."
Peter Norvig, Google Inc.

---

This book is both a reference for engineers and scientists and a teaching resource, featuring tutorial chapters and research papers on feature extraction.
"This book compiles some very promising techniques, coming from an extremely smart collection of researchers, delivering their best ideas in a competitive environment."
Trevor Hastie, Stanford University
"Feature selection is a key technology for making sense of the high dimensional data. Isabelle Guyon et al. have done a splendid job in designing a challenging competition, and collecting the lessons learned."
Bernhard Schoelkopf, Max Planck Institute
"There has been until now insufficient consideration of feature selection algorithms, no unified presentation of leading methods, and no systematic comparisons. This volume is noteworthy for the breadth of methods covered, the clarity of presentations, the unity in notation and the helpful statistical appendices."
David G. Stork, Ricoh Innovations
"Feature extraction finds application in biotechnology, industrial inspection, the Internet, radar, sonar, and speech recognition. This book will make a difference to the literature on machine learning."
Simon Haykin, Mc Master University
"This book sets a high standard as the public record of an interesting and effective competition."
Peter Norvig, Google Inc.

Content:
Front Matter....Pages I-XXIV
An Introduction to Feature Extraction....Pages 1-25
Front Matter....Pages 27-27
Learning Machines....Pages 29-64
Assessment Methods....Pages 65-88
Filter Methods....Pages 89-117
Search Strategies....Pages 119-136
Embedded Methods....Pages 137-165
Information-Theoretic Methods....Pages 167-185
Ensemble Learning....Pages 187-204
Fuzzy Neural Networks....Pages 205-233
Front Matter....Pages 235-235
Design and Analysis of the NIPS2003 Challenge....Pages 237-263
High Dimensional Classification with Bayesian Neural Networks and Dirichlet Diffusion Trees....Pages 265-296
Ensembles of Regularized Least Squares Classifiers for High-Dimensional Problems....Pages 297-313
Combining SVMs with Various Feature Selection Strategies....Pages 315-324
Feature Selection with Transductive Support Vector Machines....Pages 325-341
Variable Selection using Correlation and Single Variable Classifier Methods: Applications....Pages 343-358
Tree-Based Ensembles with Dynamic Soft Feature Selection....Pages 359-374
Margin Based Feature Selection and Infogain with Standard Classifiers....Pages 375-394
Bayesian Support Vector Machines for Feature Ranking and Selection....Pages 395-401
Nonlinear Feature Selection with the Potential Support Vector Machine....Pages 403-418
Front Matter....Pages 419-438
Combining a Filter Method with SVMs....Pages 235-235
Feature Selection via Sensitivity Analysis with Direct Kernel PLS....Pages 439-445
Information Gain, Correlation and Support Vector Machines....Pages 447-462
Mining for Complex Models Comprising Feature Selection and Classification....Pages 463-470
Combining Information-Based Supervised and Unsupervised Feature Selection....Pages 471-488
An Enhanced Selective Na?ve Bayes Method with Optimal Discretization....Pages 489-498
An Input Variable Importance Definition based on Empirical Data Probability Distribution....Pages 499-507
Front Matter....Pages 509-516
Spectral Dimensionality Reduction....Pages 517-517
Constructing Orthogonal Latent Features for Arbitrary Loss....Pages 519-550
Large Margin Principles for Feature Selection....Pages 551-583
Feature Extraction for Classification of Proteomic Mass Spectra: A Comparative Study....Pages 585-606
Sequence Motifs: Highly Predictive Features of Protein Function....Pages 607-624
Back Matter....Pages 625-645
....Pages 647-778