Personal tools

You are here: Home → Publications → Optimal Hyperplane Classifier with Adaptive Norm

News: Call for NIPS 2008 Kernel Learning Workshop Submissions 2008-09-30; Tutorials uploaded 2008-05-13; Machine Learning Summer School / Course On The Analysis On Patterns 2007-02-12; New Kernel-Machines.org server 2007-01-30; Call for participation: The 2006 kernel workshop, "10 years of kernel machines" 2006-10-06; More news…

Document Actions

K. Tsuda (1999)

Optimal Hyperplane Classifier with Adaptive Norm

ETL.

Abstract

We propose the optimal hyperplane classifier in which the norm also adapts for learning. Both hyperplane and norm parameters are optimized so that the generalization bound characterized by entropy number is minimized.The effectiveness of this classfier is validated in experiments.

URL http://www.etl.go.jp/~tsudak/tr999.ps.gz

by admin — last modified 2007-01-31 11:08