Home

About

Advanced Search

Browse by Discipline

Scientific Societies

E-print Alerts

Add E-prints

E-print Network
FAQHELPSITE MAPCONTACT US


  Advanced Search  

 
L. Rutkowski et al. (Eds.): ICAISC 2010, Part I, LNAI 6113, pp. 430-436, 2010. Springer-Verlag Berlin Heidelberg 2010
 

Summary: L. Rutkowski et al. (Eds.): ICAISC 2010, Part I, LNAI 6113, pp. 430-436, 2010.
© Springer-Verlag Berlin Heidelberg 2010
Canonical Correlation Analysis for Multiview
Semisupervised Feature Extraction
Olcay Kursun1
and Ethem Alpaydin2
1
Department of Computer Engineering, Istanbul University, 34320, Avcilar, Istanbul, Turkey
okursun@istanbul.edu.tr
2
Department of Computer Engineering, Bogazici University, 34342, Bebek, Istanbul, Turkey
alpaydin@boun.edu.tr
Abstract. Hotelling's Canonical Correlation Analysis (CCA) works with two sets of
related variables, also called views, and its goal is to find their linear projections
with maximal mutual correlation. CCA is most suitable for unsupervised feature
extraction when given two views but it has been also long known that in supervised
learning when there is only a single view of data given, the supervision signal (class-
labels) can be given to CCA as the second view and CCA simply reduces to Fisher's
Linear Discriminant Analysis (LDA). However, it is unclear how to use this
equivalence for extracting features from multiview data in semisupervised setting

  

Source: Alpaydın, Ethem - Department of Computer Engineering, Bogaziçi University

 

Collections: Computer Technologies and Information Sciences