Canonical Correlation Partial Least Squares

Uwe KRUGER, S. Joe QIN

Research output: Journal PublicationsJournal Article (refereed)peer-review

4 Citations (Scopus)

Abstract

In this paper. a novel algoritilln. referred to as Canonical Correlation Partial Least Squares or CCPLS is introduced. The core algorithm is based on the deflation procedure of partial least squares and the cost function of Canonical Correlation Analysis for detennining linear combinations of two sets of variables.
Original languageEnglish
Pages (from-to)1603-1608
Number of pages6
JournalIFAC Proceedings Volumes
Volume36
Issue number16
DOIs
Publication statusPublished - Sept 2003
Externally publishedYes
Event13th IFAC Symposium on System Identification (SYSID 2003) - , Netherlands
Duration: 27 Aug 200329 Aug 2003

Keywords

  • Correlation Coefficients
  • Covariance
  • Identification Algorithms
  • Least Squares Estimation
  • Linear Systems
  • Parameter Identification

Fingerprint

Dive into the research topics of 'Canonical Correlation Partial Least Squares'. Together they form a unique fingerprint.

Cite this