Some remarks on the functional relation between canonical correlation analysis and partial least squares. (English) Zbl 1510.62255
Summary: This paper deals with the functional relation between multivariate methods of canonical correlation analysis (CCA), partial least squares (PLS) and also their kernelized versions. Both methods are determined by the solution of the respective optimization problem, and result in algorithms using spectral or singular decomposition theories. The solution of the parameterized optimization problem, where the boundary points of a parameter give exactly the results of CCA (resp. PLS) method leads to the vector functions (paths) of eigenvalues and eigenvectors or singular values and singular vectors. Specifically, in this paper, the functional relation means the description of classes into which the given paths belong. It is shown that if input data are analytical (resp. smooth) functions of a parameter, then the vector functions are also analytical (resp. smooth). Those approaches are studied on three practical examples of European tourism data.
MSC:
| 62H20 | Measures of association (correlation, canonical correlation, etc.) |
| 62J05 | Linear regression; mixed models |
| 62H25 | Factor analysis and principal components; correspondence analysis |
Keywords:
multivariate analysis; canonical correlation analysis; partial least squares; optimization; analytical spectral decomposition; paths of eigenvalues and eigenvectors; functional relation; kernels; tourismReferences:
| [1] | Tenenhaus A, Tenenhaus M. Regularized generalized canonical correlation analysis. Psychometrika. 2011;76(2):257-284. doi: 10.1007/s11336-011-9206-8[Crossref], [Web of Science ®], [Google Scholar] · Zbl 1284.62753 |
| [2] | Martinez-Ruiz A, Aluja-Banet T. Toward the definition of a structural equation model of patent value: PLS path modelling with formative construct. REVSTAT Stat J. 2009;7(3):265-290. [Web of Science ®], [Google Scholar] |
| [3] | Wold H. Path models with latent variables: the NIPALS approach. In: Blalock HM, Aganbegian A, Borodkin FM, Boudon R, Capecchi V, editors. Quantitative sociology: international perspectives on mathematical and statistical model building. New York: Academic Press, INC.; 1975. p. 307-357. [Google Scholar] |
| [4] | Bougeard S, Hanafi M, Qannari EM. Continuum redundancy-PLS regression: a simple continuum approach. Comput Statist Data Anal. 2008;52(7):3686-3696. doi: 10.1016/j.csda.2007.12.007[Crossref], [Web of Science ®], [Google Scholar] · Zbl 1452.62489 |
| [5] | Hotelling H. The most predictable criterion. J Educ Psychol. 1935;26(2):139-142. doi: 10.1037/h0058165[Crossref], [Google Scholar] |
| [6] | Hotelling H. Relations between two sets of variables. Biometrika. 1936;28(3/4):321-377. doi: 10.2307/2333955[Crossref], [Google Scholar] · Zbl 0015.40705 |
| [7] | Wegelin JA. A survey of partial least squares (PLS) methods, with emphasis on two-block case. Seattle: University of Washington; 2000. [Google Scholar] |
| [8] | Ewerbring, LM, Luk, FT. Canonical correlation and generalized SVD: applications and new algorithms. J Comput Appl Math. 1989;27:37-52. doi: 10.1016/0377-0427(89)90360-9[Crossref], [Web of Science ®], [Google Scholar] · Zbl 0676.65147 |
| [9] | Malec L. On the multivariate processing of rank-deficient tourism data. Eur J Tour Hosp Recreat. 2013;4(3):181-203. [Google Scholar] |
| [10] | Yamamoto H, Yamaji H, Fukusaki E, Ohno H, Fukuda H. Canonical correlation analysis for multivariate regression and its application to metabolic fingerprinting. Biochem Eng J. 2008;40:199-204. doi: 10.1016/j.bej.2007.12.009[Crossref], [Web of Science ®], [Google Scholar] |
| [11] | Bílková D, Budinský P, Vohánka V. Pravděpodobnost a statistika [Probability and statistics]. Plzeň: Aleš Čeněk; 2009. [Google Scholar] |
| [12] | Fukuda K. Age-period-cohort decompositions using principal components and partial least squares. J Stat Comput Simul. 2011;81(12):1871-1878. doi: 10.1080/00949655.2010.507763[Taylor & Francis Online], [Web of Science ®], [Google Scholar] · Zbl 1365.62466 |
| [13] | Stone M, Brooks RJ. Continuum regression: cross-validated sequentially constructed prediction embracing ordinary least squares, partial least squares and principal component regression. J R Stat Soc Ser B Stat Methodol. 1990;52(2):237-269. [Web of Science ®], [Google Scholar] · Zbl 0708.62054 |
| [14] | Lee MH. Continuum direction vectors in high dimensional low sample size data [dissertation]. Chapel Hill: University of North Carolina; 2007. [Google Scholar] |
| [15] | Borga M, Landelius T, Knutsson H. A unified approach to PCA, PLS, MLR and CCA. Linkoping: Report LiTH-ISY-R-1992, ISY, SE-581 83; 1997. [Google Scholar] |
| [16] | Aydin B, Marron JS. Analyzing collaborative forecast and response networks. Stat.AP, arXiv: 1306.2062v1; 2013. [Google Scholar] |
| [17] | Lang PM, Brenchley JM, Nieves RG, Halivas JH. Cyclic subspace regression. J Multivariate Anal. 1998;65(1):58-70. doi: 10.1006/jmva.1997.1727[Crossref], [Web of Science ®], [Google Scholar] · Zbl 1005.65015 |
| [18] | Kato T. Perturbation theory for linear operators. Berlin: Springer-Verlag; 1995. [Crossref], [Google Scholar] · Zbl 0836.47009 |
| [19] | Bunse-Gerstner A, Byers R, Mehrmann V, Nichols NK. Numerical computation of an analytic singular value decomposition of a matrix valued function. Numer Math. 1991;60(1):1-39. doi: 10.1007/BF01385712[Crossref], [Web of Science ®], [Google Scholar] · Zbl 0743.65035 |
| [20] | Seber GAF. Multivariate observations. New York: Wiley & Sons; 2004. [Google Scholar] |
| [21] | Golub GH, Van Loan CF. Matrix computations. 3rd ed.Baltimore: The Johns Hopkins University Press; 1996. [Google Scholar] · Zbl 0865.65009 |
| [22] | Harville DA. Matrix algebra from a statistician’s perspective. New York: Springer-Verlag; 1997. [Crossref], [Google Scholar] · Zbl 0881.15001 |
| [23] | Malec L, Malec M. Application of two-set multivariate statistical methods to the Czech Republic arrival tourism data. In: Löster T, Pavelka T, editors. The 7th international days of statistics and economics. Proceedings; Prague; 2013. p. 937-946. [Google Scholar] |
| [24] | Bach FR, Jordan MI. Kernel independent component analysis. J Mach Learn Res. 2002;3:1-48. [Crossref], [Web of Science ®], [Google Scholar] · Zbl 1088.68689 |
| [25] | Gretton A, Herbrich R, Smola A, Bousquet O, Schölkopf B. Kernel methods for measuring independence. J Mach Learn Res. 2005;6:2075-2129. [Web of Science ®], [Google Scholar] · Zbl 1222.68208 |
| [26] | Hasan MA. Information criteria for reduced rank canonical correlation analysis. In: IEEE. International joint conference on neural networks (Vol. 3). Proceedings; Piscataway, NJ; 2004. p. 2215-2220. [Google Scholar] |
| [27] | Eurostat database [Internet]. Luxembourg: EU Statistical Office [cited 2014 Nov 2]. Available from: http://ec.europa.eu/eurostat/data/database[Google Scholar] |
| [28] | Lovaglio RG, Vittadini G. Multilevel dimensionality-reduction methods. Stat Methods Appl. 2012;22(2):183-207. doi: 10.1007/s10260-012-0215-2[Crossref], [Web of Science ®], [Google Scholar] · Zbl 1333.62160 |
| [29] | Kiráľová A, Malec L. Communication and managerial competencies as a prerequisite for employability of graduates in tourism. In: Soliman KS, editor. The 24th International Business Information Management Association conference. Proceedings, Milan; 2014. p. 477-487. [Google Scholar] |
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.
