A Leave-One-Out Bound for ν-Support Vector Regression

Authors:
Qin Ruxin;Chen Jing;Deng Naiyang;Tian Yingjie
Affiliations:
College of Science, China Agricultural University, 100083, Beijing, China;College of Science, China Agricultural University, 100083, Beijing, China;College of Science, China Agricultural University, 100083, Beijing, China;Research Center on Data Technology & Knowledge Economy, Chinese Academy of Sciences, 100080, Beijing, China
Venue:
ICCS '07 Proceedings of the 7th international conference on Computational Science, Part III: ICCS 2007
Year:
2007

Citing 4
Cited 0

Learning with Kernels: Support Vector Machines, Regularization, Optimization, and Beyond

Learning with Kernels: Support Vector Machines, Regularization, Optimization, and Beyond
Estimating the Generalization Performance of an SVM Efficiently

ICML '00 Proceedings of the Seventeenth International Conference on Machine Learning
Leave-one-out Bounds for Support Vector Regression

CIMCA '05 Proceedings of the International Conference on Computational Intelligence for Modelling, Control and Automation and International Conference on Intelligent Agents, Web Technologies and Internet Commerce Vol-2 (CIMCA-IAWTIC'06) - Volume 02
Bounds on Error Expectation for Support Vector Machines

Neural Computation

Quantified Score

Hi-index	0.00

Visualization

Abstract

An upper bound on the Leave-one-out (Loo) error forν-support vector regression (ν-SVR) ispresented. This bound is based on the geometrical concept ofspan. We can select the parameters of ν-SVR byminimizing this upper bound instead of the error itself, becausethe computation of the Loo error is extremely time consuming. Wealso can estimate the generalization performance ofν-SVR with the help of the upper bound. It is shownthat the bound presented herein provide informative and efficientapproximations of the generalization behavior based on two datasets.