Nonmetric multidimensional scaling: Neural networks versus traditional techniques

Article type: Research Article

Authors: van Wezel, M.C.^{a; *} | Kosters, W.A.^b

Affiliations: [a] Faculty of Economical Sciences, Department of Computer Science, Erasmus University, P.O. Box 1738, 3000 DR, Rotterdam, The Netherlands. E-mail: mvanwezel@few.eur.nl | [b] Leiden Institute of Advanced Computer Science (LIACS), Leiden University, P.O. Box 9512, 2300 RA Leiden, The Netherlands. E-mail: kosters@liacs.nl

Correspondence: [*] Corresponding author. Tel.: +31 10 4081341; Fax: +31 10 4089167

Abstract: In this paper we consider various methods for nonmetric multidimensional scaling. We focus on the nonmetric phase, for which we consider various alternatives: Kruskal's nonmetric phase, Guttman's nonmetric phase, monotone regression by monotone splines, and monotone regression by a monotone neural network. All methods are briefly described. We use sequential quadratic programming to estimate the weights of the neural network. An experimental comparison of the methods is given for various synthetic and real-life datasets. The monotone neural network performs comparable to the traditional methods.

Keywords: data visualization, MDS, multidimensional scaling, neural networks

DOI: 10.3233/IDA-2004-8606

Journal: Intelligent Data Analysis, vol. 8, no. 6, pp. 601-613, 2004