Military conflicts modeling by making use of Bayesian networks

Abstract

General approaches to modeling military conflicts are considered. A probabilistic and statistical model for computing probability of a military conflict in the form of Bayesian network was constructed. The complexity of modeling military conflicts is due to the high dynamics of interacting systems with many unknown or non-measurable variables and parameters. The constructed model takes into consideration incompleteness and fuzziness of the available information. The purpose of the model is not only in revealing existing links and dependencies between the processes selected, but also in forecasting the probability of future conflicts using results of the analysis of the current situation. A modeling was performed using available statistical data, a series of computational experiments was fulfilled and analysis of the results was performed. Statistical data was taken from open sources. The constructed model was used for computing the probability of internal and external conflicts for Ukraine. The obtained numerical results correspond to the actual situation.