Leave-one-out Bounds for Support Vector Regression Model Selection
Resource
Neural Computation 17,1188-1222
Journal
Neural Computation
Pages
1188-1222
Date Issued
2005
Date
2005
Author(s)
Chang, Ming-Wei
DOI
20060927122856648470
Abstract
Minimizing bounds of leave-one-out (loo) errors is an important & efficient approach
for support vector machine (SVM) model selection. Past research focuses on
their use for classification but not regression. In this article, we derive various loo
bounds for support vector regression (SVR) & discuss the difference from those for
classification. Experiments demonstrate that the proposed bounds are competitive
with Bayesian SVR for parameter selection. We also discuss the differentiability of
loo bounds.
Type
journal article
File(s)![Thumbnail Image]()
Loading...
Name
svrbound.pdf
Size
250.16 KB
Format
Adobe PDF
Checksum
(MD5):91f078f8413f6e3d98d81172bf61c8ed