Digital Criminology: Mathematical Methods of Prediction (Part 2)

The paper is devoted to the problems of digital criminology, the analysis of the methods of mathematical forecasting and the possibility of using them for crime research. The growing complexity of the task of crime counteraction determines the necessity for both the critical overhaul of the existing methods and the search for opportunities to go beyond the boundaries of the traditional methods of researching legal phenomena. The information and analytical work of law enforcement bodies based on the development of crime prevention programs has the key goal of the application of mathematical methods of crime analysis. The object of research is the complex of mathematical methods selected on the basis of their suitability for the purpose of criminological prediction. The authors single out the following methods: the modeling method, the correlation analysis, the analysis of rank correlations and conjugacy tables, discriminant analysis, regression analysis, dispersive analysis, covariance analysis, factor analysis, time series analysis, seasonal oscillation method, maximum likelihood method (in particular, its variety - least squares method), average annual growth rate calculation method, logical decision functions method, pattern recognition, calculus of variations, spectral analysis, Markov chains, algebra of logic, etc. Mathematical prediction in digital criminology consists in using the existing quantitative and qualitative parameters of criminality, calculating their mathematical dependence on time, space and other independent variables. The conducted research allowed the authors to state that the mathematical processing of criminological information makes it possible to improve the accuracy of predictions.