Predicting social tensions in individual countries based on military, political, economic, and demographic factors using machine learning methods

The purpose of the work is to increase the effectiveness of prediction using the analysis of military, political, economic, and demographic factors, as well as the preprocessing of selected predictors. Models for predicting social tensions in individual countries based on machine learning methods are presented. The results include machine learning models that predict internal armed conflict as the highest form of social tension, with the accuracy of calculated metrics provided.

Authors: I. A. Utkin, M. D. Mandritza, D. S. Nagorny

Direction: Informatics, Computer Technologies And Control

Keywords: prediction, social tension, military and political environment, machine learning, decision tree, forest of random trees, completeness, accuracy, ROC curve


View full article