Premium
An attempt for a mathematical foundation and evaluation of MACS, a method for multidimensional automatical cluster detection
Author(s) -
Von Eye A.,
Wiesing M.
Publication year - 1978
Publication title -
biometrical journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.108
H-Index - 63
eISSN - 1521-4036
pISSN - 0323-3847
DOI - 10.1002/bimj.197800005
Subject(s) - centroid , cluster (spacecraft) , ellipsoid , a priori and a posteriori , statistics , mathematics , multivariate statistics , computer science , algorithm , geometry , geodesy , geography , philosophy , programming language , epistemology
MACS, a method for the Multivariate Automatical Cluster Search, is described. It searches for local density centers by means of cuboids or ellipsoids, the diameters of which are built by the confidence intervals for the standard deviations of all predicting variables. This is done by iterative translation of the center of the neighbourhood U in direction of the centroid of his elements until a distance smaller than the one chosen a priori is reached. After this, U is varied in his diameter. The resulting values for the density and the centroid are utilized for the qualifications of U. MACS is tested by a synthetical data set and a plasmode.