Dynamic clustering for interval data based on \(L_2\) distance (Q880907)

The authors consider a \(k\)-clustering dynamic algorithm for clusterization of multidimensional intervals. It is based on the Minkowski-like \(L_2\)-type distance. Three types of variables standardization are considered, they are based on the dispersion of interval centres, dispersion of interval boundaries and the global range. Heterogeneity measures based on sums of squares are discussed. Results of simulations and application to car data are presented.