A tutorial on support vector machines for pattern recognition

DMCA

A tutorial on support vector machines for pattern recognition (1998)

Cached

Download Links

Other Repositories/Bibliography

DBLP

by Christopher J. C. Burges

Venue:	Data Mining and Knowledge Discovery
Citations:	3388 - 12 self

BibTeX

@ARTICLE{Burges98atutorial,
    author = {Christopher J. C. Burges},
    title = {A tutorial on support vector machines for pattern recognition},
    journal = {Data Mining and Knowledge Discovery},
    year = {1998},
    volume = {2},
    pages = {121--167}
}

OpenURL

Abstract

The tutorial starts with an overview of the concepts of VC dimension and structural risk minimization. We then describe linear Support Vector Machines (SVMs) for separable and non-separable data, working through a non-trivial example in detail. We describe a mechanical analogy, and discuss when SVM solutions are unique and when they are global. We describe how support vector training can be practically implemented, and discuss in detail the kernel mapping technique which is used to construct SVM solutions which are nonlinear in the data. We show how Support Vector machines can have very large (even infinite) VC dimension by computing the VC dimension for homogeneous polynomial and Gaussian radial basis function kernels. While very high VC dimension would normally bode ill for generalization performance, and while at present there exists no theory which shows that good generalization performance is guaranteed for SVMs, there are several arguments which support the observed high accuracy of SVMs, which we review. Results of some experiments which were inspired by these arguments are also presented. We give numerous examples and proofs of most of the key theorems. There is new material, and I hope that the reader will find that even old material is cast in a fresh light.

Keyphrases

support vector machine vc dimension pattern recognition svm solution high vc dimension mechanical analogy non-separable data good generalization performance kernel mapping technique structural risk minimization non-trivial example numerous example support vector training fresh light old material linear support vector machine generalization performance several argument key theorem observed high accuracy tutorial start gaussian radial basis function kernel new material

DMCA