МЕТОДИКА КЛАСТЕРИЗАЦІЇ ГРУПОВИХ ОБ’ЄКТІВ РАДІОМОНІТОРИНГУ ЗА КРИТЕРІЄМ НЕПРОТИРІЧЧЯ

Abstract

In presented article the method of radiomonitoring group objects clustering by non-contradiction criteria was evaluated. Investigated the use of algorithms of fuzzy cluster analysis by non-contradiction criteria. This is necessary to ensure the functional stability of multilevel hierarchical information classification process at radio monitoring group objects. The study was conducted using the data obtained from different types of hardware and system levels. The proposed approach allows using different types of attributes for classification: quantitative, obtained from the application of radio monitoring technical means and quality estimates, obtained from experts engaged in processing and with the experience in analytical work. Its special feature is the lack of the need to specify the number of clusters in radio monitoring facilities as a priori information for the use of known fuzzy clustering algorithms.Approbation of the developed technique of information process management was conducted in Matlab media, based on radiomonitoring data, obtained during the operational activities in a particular area.Based on the calculations, described above, it is concluded that the use of fuzzy clustering algorithms with the non-contradiction criteria for determination of the number of clusters on the stage of radiomonitoring objects classification provides the functional resistance of the information process under uncertainty.