Abstract:

In multidimensional scaling (MDS) carried out on the basis of a metric data matrix (interval, ratio) or interval-valued data table three approaches can be distinguished: classic-to-classic – for metric data, symbolic-to-classic) and symbolic-to-symbolic – for interval-valued data). The article presents the mdsOpt package which helps to solve the problem of choosing the optimal MDS procedure. It uses two criteria for selecting the optimal MDS procedure: Kruskal's  fit measure ( in case of symbolic-to-symbolic approach) and Hirschman-Herfindahl  index calculated based on Stress per point (Interval stress per box in case of symbolic-to-symbolic approach) values. In first part three possible approaches are characterized with theoretical background of used methods and the relationships between mdsOpt package and existing R packages. Second part explains procedure and criteria for selection of the optimal MDS procedure for metric and interval-valued data. The last contains in details the usage of package R and applications on real data sets.