Multiplicative Updatings for Support Vector Learning
Nello Cristianini, Colin Campbell, John S. Shawe-Taylor · 1999
Support Vector machines find maximal margin hyperplanes in a high dimensional feature space. Theoretical results exist which guarantee a high generalization performance when the margin is large or when the number of support vectors is small. Multiplicative-Updating algorithms are a new tool for perceptron learning whose theoretical properties are well studied. In this work we present a Multiplicative-Updating algorithm for learning Support Vector machines which exploits the particular structure of highgeneralization hypotheses, by achieving fast rate of convergence just in those situations where high generalization can be obtained, namely small number of support vectors or large margin. Keywords: Theory, support vector machines 1 Introduction Support Vector (SV) machines are a class of algorithms introduced by Vapnik and coworkers [5] for implementing nonlinear decision rules in terms of hyperplanes in high-dimensional feature spaces. Multiplicative-Updating algorithms, are a relativ...