UDC 004.855.5
DEVELOPMENT OF TWO-LEVEL CLASSIFIER FOR ELABORATE MULTIDIMENSIONAL DATA OF LARGE CAPACITIES
L. A. Demidova, PhD (technical sciences), full professor, RSREU, Ryazan; This email address is being protected from spambots. You need JavaScript enabled to view it.
Yu. S. Sokolova, lecturer, RSREU, This email address is being protected from spambots. You need JavaScript enabled to view it.
The problem of classification of elaborate multidimensional data of large capacities which is inherent in various social and economic systems has been considered. The purpose of work is the increase of classification accuracy of elaborate multidimensional data of large capacities by means of development of two-level classifier on the base of SVM algorithm at the acceptable time expenditures. At the first level of two-level classifier the group of SVM classifiers for which the training sets are formed on the base of initial data set is created. At the second level of two-level classifier SVM classifier with the use of the modified particle swarm algorithm for which the training set is formed on the base of the set of support vectors received at the first level of two-level classifier is developed. The given results of experimental studies confirm application efficiency of the offered two-level classifier for the problem of classification of elaborate multidimensional data of large capacities.
Keywords: SVM-classifier, support vectors, kernel function type, kernel function parameters, regularization parameter, particle swarm algorithm.