Please use this identifier to cite or link to this item:
Scopus Web of Science® Altmetric
Type: Journal article
Title: UBoost: Boosting with the Universum
Other Titles: {cal U}Boost: Boosting with the Universum
Author: Shen, C.
Wang, P.
Shen, F.
Wang, H.
Citation: IEEE Transactions on Pattern Analysis and Machine Intelligence, 2012; 34(4):825-832
Publisher: IEEE Computer Soc
Issue Date: 2012
ISSN: 0162-8828
Statement of
Chunhua Shen, Peng Wang, Fumin Shen and Hanzi Wang
Abstract: It has been shown that the Universum data, which do not belong to either class of the classification problem of interest, may contain useful prior domain knowledge for training a classifier [1], [2]. In this work, we design a novel boosting algorithm that takes advantage of the available Universum data, hence the name UBoost. UBoost is a boosting implementation of Vapnik’s alternative capacity concept to the large margin approach. In addition to the standard regularization term, UBoost also controls the learned model’s capacity by maximizing the number of observed contradictions. Our experiments demonstrate that UBoost can deliver improved classification accuracy over standard boosting algorithms that use labeled data alone.
Keywords: Universum; kernel methods; boosting; column generation; convex optimization
Rights: Copyright 2012 IEEE
RMID: 0020116765
DOI: 10.1109/TPAMI.2011.240
Appears in Collections:Computer Science publications

Files in This Item:
File Description SizeFormat 
hdl_67027.pdfAccepted version481.95 kBAdobe PDFView/Open

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.