Issue 9 - Issue 9
Error analysis for support vector machine classifiers on unite sphere of Euclidean space
Abstract
We provide the error analysis of p-norm Support Vector Machine (SVM) classifiers on unit sphere of n-dimensional Euclidean space. The approximation error on the unite sphere are estimated with spherical harmonics approximation. We also provide the standard estimation of the sample error, and derive the explicit learning rate.
Paper Details
PaperID: 80052818243
Author's Name: Bao, H., Ye, P.
Volume: Issue 9
Issues: Issue 9
Keywords: Ces̀aro means, Learning rate, Reproducing kernel hilbert spaces, Spherical harmonics, Support vector machine classification
Year: 2011
Month: September
Pages: 3023 - 3030
