Modeling Military Conflicts Using Bayesian Networks

In this investigation, general approaches to military conflicts modeling are analyzed. The research focuses on a probabilistic statistical model used to determine the probability of military conflict eruption in the form of a Bayesian network. The complexity of military conflict modeling is caused by a large-scale dynamics of interacting systems with a large number of unknown or immeasurable variables and parameters. The resulting model incorporates the incompleteness and fuzziness of the available information. Consequently, the model is aimed at revealing the existing relationships and dependencies between the selected processes as well as the future conflicts probability forecasting based on the current situation analysis. Furthermore, the research presents the modeling procedure with the use of the available statistical data, describes a series of computing experiments, and analyses the results. Statistical data used for sampling was taken from public sources. The created model was used to measure the probability of internal and external conflicts in Ukraine. The obtained numerical results correspond to the current situation.