System for prediction of offenses in the field of public security

In this article, questions of the theory of prediction of offenses (predictive analytics) are considered. The authors investigated the question of modern technology allowing to move from theoretical or simple simplifying solutions to complex mathematical models and use productive tools for processing large data sets, including using artificial intelligence. The experience of foreign countries that have achieved success in the field of applications of technologies of "neural network algorithms" (adaptive system for prediction of offenses and crimes), for solving problems of public security was analyzed. The project also envisages the creation one of the first in Russia specialized adaptive system for prediction of crime and support for decision-making in the sphere of public security "Pretorian". Thus, the authors come to the conclusion that the importance of large data processing tools, predictive analysts, will grow in the next decades, which form the basis for effective methods of work to improve the effectiveness of public security.