Evolutionary strategies to avoid local minima in multidimensional scaling. (English) Zbl 05280175

Etschberger, Stefan; Hilbert, Andreas

Schader, Martin (ed.) et al., Between data science and applied data analysis. Proceedings of the 26th annual conference of the Gesellschaft für Klassifikation e. V., Mannheim, Germany, July 22–24, 2002. Berlin: Springer. Stud. Classification Data Anal. Knowl. Organ., 209-217 (2003).

Summary: Multidimensional scaling is very widely used in exploratory data analysis. It is mainly used to represent sets of objects with respect to their proximities in a low dimensional Euclidean space. Widely used optimization algorithms try to improve the representation via shifting its coordinates in direction of the negative gradient of a corresponding fit function. Depending on the initial configuration, the chosen algorithm and its parameter settings there is a possibility for the algorithm to terminate in a local minimum. This article describes the combination of an evolutionary model with a non-metric gradient solution method to avoid this problem. Furthermore a simulation study compares the results of the evolutionary approach with one classic solution method.
For the entire collection see [Zbl 1023.00021].

MSC:

62-XX

Statistics