A multi-view two-level classification method for generalized multi-instance problems

From National Research Council Canada

DOIResolve DOI: https://doi.org/10.1109/BigData.2014.7004363
AuthorSearch for: ; Search for: ; Search for: ; Search for: ; Search for: 1
Name affiliation
  1. National Research Council Canada. Information and Communication Technologies
FormatText
TypeArticle
Proceedings title2014 IEEE International Conference on Big Data (Big Data)
Conference2nd IEEE International Conference on Big Data, IEEE Big Data 2014, October 27-30, 2014, Washington, DC, USA
ISBN978-1-4799-5665-4
978-1-4799-5666-1
Article number7004363
Pages104–111; # of pages: 8
SubjectBig data; Unsupervised learning; Classification framework; Classification methods; Empirical studies; Learning methods; Multi-instance problems; Multi-view learning; Multi-views; Supervised and unsupervised learning; Learning systems
Abstract
Multi-instance (MI) learning is different than standard propositional classification, as it uses a set of bags containing many instances as input. While the instances in each bag are not labeled, the bags themselves are, as positive or negative. In this paper, we present a novel multi-view, two-level classification framework to address the generalized multi-instance problems. We first apply supervised and unsupervised learning methods to transform a MI dataset into a multi-view, single meta-instance dataset. Then we develop a multi-view learning approach that can integrate the information acquired by individual view learners on the meta-instance dataset from the previous step, and construct a final model. Our empirical studies show that the proposed method performs well compared to other popular MI learning methods.
Publication date
PublisherIEEE
LanguageEnglish
Peer reviewedYes
NPARC number21275640
Export citationExport as RIS
Report a correctionReport a correction
Record identifier85e659c5-7de8-4d10-a9fa-56eb8c25f143
Record created2015-07-14
Record modified2016-08-18
Report a problem on this page
Date modified: