Investigation of effects on multidimensional scaling algorithms of distance functions and applications

Multidimensional scaling (MDS) is a statistical tool for constructing a low-dimensional configuration to represent the relationships between the objects. In this study, the effects of distance functions on the stress will be examined in metric multidimensional scaling algorithms. To study the effects on stress value of distance measures used in both simulation and real data sets. The distance meauserment having the best performance is Euclidean distance according to stress value. Also in this study, two applications are made to demonstrate the pratical applicability of metric MDS algorithms. In first application, the first-level statistical region units will be compared in terms of socio-economic indicators by metric multidimensional scalling algorithms. In second application, it is aimed to examine the position of European states and Turkey and display possible similarities or dissimilarities between them, from the point of their socio-economic indicators.