Support vector machines (SVMs) perform pattern recognition between two point classes by finding a decision surface determined by certain points of the training set, termed support vectors (SV). This surface, which in some feature space of possibly infinite dimension can be regarded as a hyperplane, is obtained from the solution of a problem of quadratic programming that depends on a regularization parameter. In this article, we study some mathematical properties of support vectors and show that the decision surface can be written as the sum of two orthogonal terms, the first depending on only the margin vectors (which are SVs lying on the margin), the second proportional to the regularization parameter. For almost all values of the parameter, this enables us to predict how the decision surface varies for small parameter changes. In the special but important case of feature space of finite dimension m, we also show that there are at most m + 1 margin vectors and observe that m + 1 SVs are usually sufficient to determine the decision surface fully. For relatively small m, this latter result leads to a consistent reduction of the SV number.
[1]
R. Courant,et al.
Methods of Mathematical Physics
,
1962
.
[2]
Vladimir Vapnik,et al.
Chervonenkis: On the uniform convergence of relative frequencies of events to their probabilities
,
1971
.
[3]
M. S. Bazaraa,et al.
Nonlinear Programming
,
1979
.
[4]
Vladimir Vapnik,et al.
Estimation of Dependences Based on Empirical Data: Springer Series in Statistics (Springer Series in Statistics)
,
1982
.
[5]
Bernhard Schölkopf,et al.
Comparison of View-Based Object Recognition Algorithms Using Realistic 3D Models
,
1996,
ICANN.
[6]
Federico Girosi,et al.
Training support vector machines: an application to face detection
,
1997,
Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition.
[7]
Bernhard Schölkopf,et al.
Comparing support vector machines with Gaussian kernels to radial basis function classifiers
,
1997,
IEEE Trans. Signal Process..
[8]
Vladimir N. Vapnik,et al.
The Nature of Statistical Learning Theory
,
2000,
Statistics for Engineering and Information Science.