Contemporary Mathematics 2007; 226 pp; softcover Volume: 443 ISBN10: 0821841955 ISBN13: 9780821841952 List Price: US$71 Member Price: US$56.80 Order Code: CONM/443
 These proceedings feature some of the latest important results about machine learning based on methods originated in Computer Science and Statistics. In addition to papers discussing theoretical analysis of the performance of procedures for classification and prediction, the papers in this book cover novel versions of Support Vector Machines (SVM), Principal Component methods, Lasso prediction models, and Boosting and Clustering. Also included are applications such as multilevel spatial models for diagnosis of eye disease, hyperclique methods for identifying protein interactions, robust SVM models for detection of fraudulent banking transactions, etc. This book should be of interest to researchers who want to learn about the various new directions that the field is taking, to graduate students who want to find a useful and exciting topic for their research or learn the latest techniques for conducting comparative studies, and to engineers and scientists who want to see examples of how to modify the basic highdimensional methods to apply to real world applications with special conditions and constraints. Readership Research mathematicians interested in machine learning. Table of Contents  J. S. Verducci and X. Shen  Introduction
 J. Wang, X. Shen, and W. Pan  On transductive support vector machines
 X. Deng, M. Yuan, and A. Sudjianto  A note on robust kernel principal component analysis
 Y. Liu, H. H. Zhang, C. Park, and J. Ahn  The \(L_q\) support vector machine
 Y. Wu and Y. Liu  On multicategory truncatedhingeloss support vector machines
 A. B. Owen  A robust hybrid of lasso and ridge regression
 Y. Kim, Y. Kim, and J. Kim  A gradient descent algorithm for LASSO
 B. Li and P. K. Goel  Additive regression trees and smoothing splinespredictive modeling and interpretation in data mining
 E. P. Fokoué  Estimation of atom prevalence for optimal prediction
 C. Rudin, R. E. Schapire, and I. Daubechies  Precise statements of convergence for AdaBoost and arcgv
 K. Marsolo, S. Parthasarathy, M. Twa, and M. Bullimore  Ensemblelearning by modelbased spatial averaging
 H. Zou, J. Zhu, S. Rosset, and T. Hastie  Automatic bias correction methods in semisupervised learning
 S. Wang and J. Zhu  Variable selection for modelbased highdimensional clustering
 W. Pan and X. Shen  Semisupervised learning via constraints
 M. Steinbach, P  N. Tan, H. Xiong, and V. Kumar  Objective measures for association pattern analysis
