Support vector machines and kernel methods have proven as powerful tools in numerous applications and have thus gained widespread popularity e.g. in machine learning and bioinformatics. They certainly do not violate the "no free lunch" theorem, i.e. they do not outperform every other technique in each and every pattern recognition application. But used properly they certainly constitute a very handy and powerful tool in the data analysis and pattern recognition toolkit.

This page is intended as an easy access point to this vast field with its own language and conventions. Thus, for the sake of simplicity and clarity it only covers a small and subjective collection of introductory links.

Tutorials/Textbooks

• A Tutorial on Support Vector Machines for Pattern Recognition by Chris Burges
• Support Vector Machines for Classification and Regression by Steve Gunn with accompanying Matlab toolbox (Caveat: Uses very inefficient optimization, not recommendable for use in real applications.)
• A tutorial on ?-Support Vector Machines
• An Introduction to Support Vector Machines by Nello Cristianini and John Shawe-Taylor
• Learning with Kernels by Bernhard Schölkopf and Alex Smola
• Support Vector Learning Ph.D. by Bernhard Schölkopf
• Kernel Methods for Pattern Analysis by J. Shawe-Taylor and N. Cristianini
• Pattern Classification by R.O. Duda, P.E. Hart and D.G. Stork (not dedicated to SVMs but a "classic" textbook on pattern recognition)
• The Elements of Statistical Learning by T. Hastie, R. Tibshirani and J. Friedman (another "classic" general textbook)
• Pattern Recognition and Machine Learning by C.M. Bishop (very educative general textbook)
• A flexible classification approach with optimal generalization performance: support vector machines by A.I. Belousov, S.A. Verzakov and J. von Frese

Websites

• support-vector-machines.org
• kernel-machines.org
• Journal of Machine Learning Research

Patents

• Note that SVMs are subject to patent coverage. In case of commercial applications you might want to consider patents held e.g. by Health Discovery Corporation

Software

• SVM Optimizer, DA-SOL's simple but effective Matlab GUI for developing SVM classifiers
• Spider, very comprehensive Matlab implementation from J. Weston et al.
• Shogun, machine-learning toolbox with interface to Matlab
• SVMlight by T. Joachims
• LIBSVM, the engine from Chih-Chung Chang and Chih-Jen Lin for most other packages and commercial softwares
• perClass, a comprehensive Pattern Recognition suite
• PLS-Toolbox, the Matlab chemometrics toolbox
• Aspentech Unscrambler, multivariate analysis software
• Relevance Vector Machine
• Least Squares-SVM
• SVM in R
  Kernlab
  Interface to LIBSVM
• Software overview from kernel-machines.org
• PRTools (not dedicated to SVMs but a very comprehensive Matlab toolbox for pattern recognition)

Reviews

• An introduction to Kernel-Based Learning Algorithms by K.-R. Müller et al.
• Statistical Pattern Recognition: A Review by A.K. Jain, R.P.W. Duin and J. Mao

Miscellanous Topics

• Feature selection by I. Guyon and A. Elisseeff
• Feature Extraction, Foundations and Applications, book ed. by I. Guyon et al.
• One-Class SVM by D. Tax and R.P.W. Duin
• Kernel-PCA by B. Schölkopf, A. Smola and K.-R. Müller (see also Support Vector Learning)
• Eigenproblems (Kernel-PCA, CCA, PLS, Fisher Discriminant Analysis)
• Data fusion by G.R.G. Lanckriet et al.
• Multiple Kernel Learning by S. Sonnenburg et al.
• Relevance Vector Machine by M. Tipping

Some Chemometric Applications

• SVM Application List (applications in general)
• Applications of Support Vector Machines in Chemistry
• Automatic masking in multivariate image analysis using support vector machines by J. Jay Liu et al.
• Combination of support vector machines (SVM) and near-infrared (NIR) imaging spectroscopy for the detection of meat and bone meal (MBM) in compound feeds by J. A. Fernández Pierna et al.
• Support Vector Machine Regression in Chemometrics by K. Bennett et al.
• Comparing support vector machines to PLS for spectral regression applications by U. Thissen et al.
• Support Vector Machines for the Discrimination of Analytical Chemical Data: Application to the Determination of Tablet Production by Pyrolysis-gas Chromatography Mass Spectrometry by S. Zomer et al.
• Applicational aspects of support vector machines by A.I. Belousov, S.A. Verzakov and J. von Frese

Dr. Juergen von Frese
Data Analysis Solutions S.L.
Telephone: +34 871 811 605
E-Mail: jvf@da-sol.com