|
NeuroCOLT
Technical Report NC-TR-98-016
Multiplicative
Updatings for Support-Vector Learning
Nello Cristianini
& Colin Campbell
Dept. of Engineering Mathematics
Bristol
John Shawe-Taylor
Dept. of Computer Science
RHUL
Keywords:
perceptron, decision tree, maximal margin
Received:
04-JUN-98
Abstract
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 high-generalization
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.
Download Compressed Postscript
|