# Maximization of sums of quotients of quadratic forms and some generalizations

- 191 Downloads
- 10 Citations

## Abstract

Monotonically convergent algorithms are described for maximizing six (constrained) functions of vectors x, or matrices*X* with columns x_{1}, ..., x_{ r }. These functions are h_{1}(x)=Σ_{ k } (x′*A*_{k}x)(x′*C*_{k}x)^{−1}, H_{1}(*X*)=Σ_{ k } tr (*X′A*_{k}*X*)(*X′C*_{k}*X*)^{−1}, h_{1}(*X*)=Σ_{ k } Σ_{ l } (x′_{ l }*A*_{k}x_{ l }) (x′_{ l }*C*_{k}x_{ l })^{−1} with*X* constrained to be columnwise orthonormal, h_{2}(x)=Σ_{ k } (x′*A*_{k}x)^{2}(x′*C*_{k}x)^{−1} subject to x′x=1, H_{2}(*X*)=Σ_{ k } tr*(X′A*_{k}X)(X′A_{k}X)′(X′C_{k}X)^{−1} subject to*X′X*=*I*, and h_{2}(*X*)=Σ_{ k } Σ_{ l } (x′_{ l }*A*_{k}x_{ l })^{2} (x′_{ l }*C*_{k}X_{ l })^{−1} subject to*X′X*=*I*. In these functions the matrices*C*_{k} are assumed to be positive definite. The matrices*A*_{k} can be arbitrary square matrices. The general formulation of the functions and the algorithms allows for application of the algorithms in various problems that arise in multivariate analysis. Several applications of the general algorithms are given. Specifically, algorithms are given for reciprocal principal components analysis, binormamin rotation, generalized discriminant analysis, variants of generalized principal components analysis, simple structure rotation for one of the latter variants, and set component analysis. For most of these methods the algorithms appear to be new, for the others the existing algorithms turn out to be special cases of the newly derived general algorithms.

## Key words

generalized principal components analysis generalized discriminant analysis binormamin simple structure rotation## Preview

Unable to display preview. Download preview PDF.

## References

- de Leeuw, J. (1982). Generalized eigenvalue problems with positive semi-definite matrices.
*Psychometrika, 47*, 87–93.Google Scholar - de Leeuw, J. & Heiser, W. (1980). Multidimensional scaling with restrictions on the configuration. In P. R. Krishnaiah (Ed.),
*Multivariate analysis V*(pp. 501–522). Amsterdam: North Holland.Google Scholar - Dinkelbach, W. (1967). On nonlinear fractional programming.
*Management Science, 13*, 492–498.Google Scholar - Harman, H. H. (1976).
*Modern factor analysis*(3rd ed.). Chicago: University of Chicago Press.Google Scholar - Henderson, H. V., & Searle, S. R. (1981). The Vec-permutation matrix, the Vec operator and Kronecker products: A review.
*Linear and multilinear algebra, 9*, 271–288.Google Scholar - Kaiser, H. F., & Dickman, K. W. (1959).
*Analytic determination of common factors*. Unpublished manuscript, University of Illinois.Google Scholar - Kiers, H. A. L. (1990). Majorization as a tool for optimizing a class of matrix functions.
*Psychometrika, 55*, 417–428.Google Scholar - Kiers, H. A. L. (1991). Simple structure in components analysis techniques for mixtures of qualitative and quantitative variables.
*Psychometrika, 56*, 197–212.Google Scholar - Kiers, H. A. L., & ten Berge, J. M. F. (1989). Alternating least squares algorithms for simultaneous components analysis with equal component weight matrices for all populations.
*Psychometrika, 54*, 467–473.Google Scholar - Kiers, H. A. L., & ten Berge, J. M. F. (1992). Minimization of a class of matrix trace functions by means of refined majorization.
*Psychometrika, 57*, 371–382.Google Scholar - Kiers, H. A. L., & ten Berge, J. M. F. (1994). Hierarchical relations between methods for simultaneous component analysis and a technique for rotation to a simultaneous structure.
*British Journal of Mathematical and Statistical Psychology*,**47**, 109–126.Google Scholar - Kroonenberg, P. M., & de Leeuw, J. (1980). Principal component analysis of three-mode data by means of alternating least squares algorithms.
*Psychometrika, 45*, 69–97.CrossRefGoogle Scholar - McDonald, R. P. (1968). A unified treatment of the weighting problem.
*Psychometrika, 33*, 351–384.CrossRefPubMedGoogle Scholar - McDonald, R. P., Torii, Y., & Nishisato, S. (1979). Some results on proper eigenvalues and eigenvectors with applications to scaling.
*Psychometrika, 44*, 211–227.CrossRefGoogle Scholar - Meulman, J. J. (1986).
*A distance approach to nonlinear multivariate analysis*. Leiden: DSWO Press.Google Scholar - Millsap, R. E., & Meredith, W. (1988). Component analysis in cross-sectional and longitudinal data.
*Psychometrika, 53*, 123–134.CrossRefGoogle Scholar - Nierop, A. F. M. (1993).
*Multidimensional analysis of grouped variables: An integrated approach*. Leiden: DSWO Press.Google Scholar - Ostrowski, A. M. (1969).
*Solutions of equations and systems of equations*. New York: Academic Press.Google Scholar - Tatsuoka, M. M. (1971).
*Multivariate analysis: Techniques for educational and psychological research*. New York: Wiley.Google Scholar - ten Berge, J. M. F. (1983). A generalization of Kristof's theorem on the trace of certain matrix products.
*Psychometrika, 48*, 519–523.Google Scholar - ten Berge, J. M. F., Knol, D. L., & Kiers, H. A. L. (1988). A treatment of the orthomax rotation family in terms of diagonalization, and a re-examination of a singular value approach to varimax rotation.
*Computational Statistics Quarterly, 3*, 207–217.Google Scholar