Support vector regression for right censored data

Yair Goldberg, Michael R. Kosorok

Research output: Contribution to journalArticlepeer-review

Abstract

We develop a unified approach for classification and regression support vector machines for when the responses are subject to right censoring. We provide finite sample bounds on the generalization error of the algorithm, prove risk consistency for a wide class of probability measures, and study the associated learning rates. We apply the general methodology to estimation of the (truncated) mean, median, quantiles, and for classification problems. We present a simulation study that demonstrates the performance of the proposed approach.

Original languageEnglish
Pages (from-to)532-569
Number of pages38
JournalElectronic Journal of Statistics
Volume11
Issue number1
DOIs
StatePublished - 2017

Bibliographical note

Funding Information:
The authors thank Danyu Lin for many helpful discussions and suggestions. The first author was funded in part by a Gillings Innovation Laboratory (GIL) award at the UNC Gillings School of Global Public Health and by Norsk SykepleierforbundNSF grant DMS-1407732. The second author was funded in part by NCINational Cancer Institute grant CA142538 and by NSFNorsk Sykepleierforbund grant DMS-1407732.

Publisher Copyright:
© 2017, Institute of Mathematical Statistics. All rights reserved.

Keywords

  • Generalization error
  • Misspecification models
  • Right censored data
  • Support vector regression
  • Universal consistency

ASJC Scopus subject areas

  • Statistics and Probability
  • Statistics, Probability and Uncertainty

Fingerprint

Dive into the research topics of 'Support vector regression for right censored data'. Together they form a unique fingerprint.

Cite this