DEVELOPMENT OF A UNIVERSAL METHOD FOR PREDICTING THE BEHAVIOR OF OBJECTS IN AN INFORMATION SYSTEM: ANALYSIS OF CURRENT DEVELOPMENTS IN THE FIELD OF BEHAVIORAL ANALYTICS AND MATHEMATICAL THEORIES

Mokhorev Dmitrii Evgenevich

The study considers the possibility of creating a universal method for predicting the behavior of objects in an information system. To this end, an analysis of current developments in the field of behavioral analytics is carried out and the main mathematical theories underlying them are highlighted. The criteria of universality of algorithms for analyzing the behavior of objects are presented, which make it possible to assess the possibilities and limitations of their use in various conditions and software environments. As a result of the analysis, the problems preventing the creation of a universal method for predicting the behavior of objects in an information system are formulated.

Keywords:information security, machine learning, behavioral analysis, UEBA, Bayes theorem, time series analysis, fuzzy clustering method.