Abstract:

In this study, a new aggregation algorithm was developed for the accumulation of many variable homogeneous and heterogeneous big data. The fragmentation in the heterogeneous data determines the number and location of the sets. The number of fragmentations in heterogeneous data is determined based on both graphic and calculated methods. In graphic methods, the probability graphs of each variable are used, and in calculated methods, the single variable normal mix distribution of each variable is used. Genetic algorithms are used to determine the location and structure of accumulation centers that correspond to the fragmentation in heterogeneous data. Models determined on the basis of the number and structure of the accumulation centers are obtained using normal mixed distribution. Each cement center in the normal karma models responds to the fragmentation in the variables.  The best karma model that fits the data structure from the normal karma models is obtained using the information criteria obtained from the normal karma distribution.

Keywords: