On the Non-Existence of Optimal Solutions and the Occurrence of “Degeneracy” in the CANDECOMP/PARAFAC Model
Authors:
Open Access
- 1 September 2008
- journal article
- Published by Cambridge University Press (CUP) in Psychometrika
- Vol. 73 (3) , 431-439
Abstract
The CANDECOMP/PARAFAC (CP) model decomposes a three-way array into a prespecified number of R factors and a residual array by minimizing the sum of squares of the latter. It is well known that an optimal solution for CP need not exist. We show that if an optimal CP solution does not exist, then any sequence of CP factors monotonically decreasing the CP criterion value to its infimum will exhibit the features of a so-called “degeneracy”. That is, the parameter matrices become nearly rank deficient and the Euclidean norm of some factors tends to infinity. We also show that the CP criterion function does attain its infimum if one of the parameter matrices is constrained to be column-wise orthonormal.Keywords
This publication has 29 references indexed in Scilit:
- Convergence of the sequence of parameters generated by alternating least squares algorithmsComputational Statistics & Data Analysis, 2006
- Tensorial extensions of independent component analysis for multisubject FMRI analysisNeuroImage, 2005
- Structure-seeking multilinear methods for the analysis of fMRI dataNeuroImage, 2004
- Three-way (PARAFAC) factor analysis: examination and comparison of alternative computational methods as applied to ill-conditioned dataChemometrics and Intelligent Laboratory Systems, 1998
- A fast non-negativity-constrained least squares algorithmJournal of Chemometrics, 1997
- Simple Structure in Component Analysis Techniques for Mixtures of Qualitative and Quantitative VariablesPsychometrika, 1991
- An Alternating Least Squares Algorithm for Fitting the Two- and Three-Way Dedicom Model and the Idioscal ModelPsychometrika, 1989
- Three-Way Metric Unfolding Via Alternating Weighted Least SquaresPsychometrika, 1985
- Factor analysis of tongue shapesThe Journal of the Acoustical Society of America, 1977
- Analysis of Individual Differences in Multidimensional Scaling Via an N-way Generalization of “Eckart-Young” DecompositionPsychometrika, 1970