Scientific Publications with referee

A bridge between Tucker-1 and generalised canonical correlation.

 

En gruppe metoder for analyse av multiple data matriser betraktes. Spesielt studeres relasjonen mellom to viktige metoder av denne typen, GCA og Tucker-1. Det vises at disse to kan settes på en felles formel som også åpner opp for en serie mellomløsninger eller kompromiss. Dette kan være viktig for valg av fornuftig metode til analyse. Metodene illustreres med data fra sensorisk analyse av pølser.

Facts

Year 2005
Abstract This paper concerns tools for analyzing relationships between and within multiple data matrices. A new unified approch is developed, bridging two existing methods; Carrol's Generalized Canonical Analysis (GCA) and the Tucker-1 metod for principal component analysis of multiple matrices. GCA and Tucker-1 are shown to correspond to particular choices of a ridge parameter. The unified method may again be generalized to a larger space of methods.
Reference Dahl, T., Næs, T. 2006. A bridge between Tucker-1 and generalised canonical correlation. Computational Statistics and Data Analysis, Vol 50, No 11, pp 3086-3098.
Publisher Computational Statistics and Data Analysis,

Related persons

  • Tormod Næs

    Senior Research Scientist

    Phone: +47 64970165

    Cellphone: +47 913 52 032