Pages: 101-111
Published: 30.12.2017
Abstract: The article describes the process of modeling block-diagonal correlation matrices and corresponding multidimensional -samples using no more than three methods. One of the methods used is new. Its application was carried out with the use of the new Optimization Problem No. 4 (problems No. 1,2,3 were solved in [10,11]). The dimension n of the correlation matrices is n=n1+2k+ 1, where n1=3,4, ... 10, k = 1, ..., 10. In the case n1=0, the cases n=2k+1 and n =2k are considered. An example of calculations for n=6 is given. Multidimensional -samples can be applied in multidimensional classification problems and cluster analysis problems.
Key words: inverse model jf the principal component analysis, sample
|
