Support vector machines regression with unbounded sampling. Issue 9 (4th July 2019)

• Record Type: Journal Article
• Authors: Tong, Hongzhi; Chen, Di-Rong; Yang, Fenghong
• Is Part Of: Applicable analysis. Volume 98:Issue 9(2019)
• Page Start: 1626
• Page End: 1635
• Publication Date: 2019-07-04
• Subjects: Support vector machines regression -- unbounded sampling -- Hoeffding inequality -- learning rate; 68T05 -- 62J02; Mathematical analysis -- Periodicals; 515
• DOI: 10.1080/00036811.2018.1437416
• Languages: English
• ISSNs: 0003-6811

Abstract

ABSTRACT: Uniform boundedness of output variables is a standard assumption in most theoretical analysis of regression algorithms. This standard assumption has recently been weaken to a moment hypothesis in least square regression (LSR) setting. Although there has been a large literature on error analysis for LSR under the moment hypothesis, very little is known about the statistical properties of support vector machines regression with unbounded sampling. In this paper, we fill the gap in the literature. Without any restriction on the boundedness of the output sampling, we establish an ad hoc convergence analysis for support vector machines regression under very mild conditions.