Support vector machine

A support vector machine (SVM) is an example of a fast, lightweight machine learning technique first discussed by Vladimir Vapnik. Support vector machines are often used as non-linear classifiers, although they can always be shown to be equivalent to linear classifiers of hyperplanes in a high-dimensional space.

An SVM uses a kernel function (often non-linear) to transform distances between sample points before making comparisons. Even in the non-linear case, it can be shown that this is equivalent to performing a transformation of the classification space into another, higher dimensional, space and using a linear classifier on this new space. This technique can also be used when the higher-dimensional space is a Hilbert space, and has an infinite number of dimensions.

References

• N. Cristianini and J. Shawe-Taylor. AN INTRODUCTION TO SUPPORT VECTOR MACHINES (and other kernel-based learning methods). Cambridge University Press, 2000. ISBN 0-521-78019-5

External links

• http://www.kernel-machines.org/
• K.-R. Müller, S. Mika, G. Rätsch, K. Tsuda, and B. Schölkopf. An introduction to kernel-based learning algorithms. (http://mlg.anu.edu.au/~raetsch/ps/review.pdf) IEEE Neural Networks, 12(2):181-201, May 2001.
• Text Categorization with Support Vector Machines: Learning with Many Relevant Features (http://citeseer.nj.nec.com/joachims98text.html)
• Look at libsvm (http://www.csie.ntu.edu.tw/~cjlin/libsvm/) by Chih-Chung Chang and Chih-Jen Lin
• The Formulation of Support Vector Machines (http://svr-www.eng.cam.ac.uk/~kkc21/thesis_main/node8.html)