PREDICTION OF INTERSTATE ARMED CONFLICTS IN WEST AFRICA USING ARTIFICIAL INTELLIGENCE TECHNIQUES.
- National School of Applied Economics and Management , University of Abomey-Calavi, 03 BP 1079, Cotonou, Benin.
- African School of Economics, 02 BP 372, Abomey-Calavi, Benin.
- National University of Sciences, Technologies, Engineering and Mathematics, BP 486, Abomey, Benin
- Abstract
- Cite This Article as
- Corresponding Author
In the paper, we tested three machine learning models to predict the occurrence of inter-state armed conflicts in West Africa. The data used was collected from various sources over the period from 1981 to 2018. The modeling was done using Python. Among the three developed models, it turned out that the random forest model was the most suitable for this prediction. The modeling revealed that two major categories of variables are the most relevant predictors: whether or not countries share borders and the difference in democracy levels. Using the results of this prediction, we also identified the risk of conflicts between countries, such as Guinea vs. Sierra Leone, Burkina Faso vs. Benin, and Ivory Coast vs. Burkina Faso. This research deepens our understanding of state-to-state conflict dynamics in the West African region. However, it has limitations partly due to the lack of dynamic data availability.
[Maurice Comlan, Horace Agossadou, Sardou Agbangla and Patrick Sotindjo (2025); PREDICTION OF INTERSTATE ARMED CONFLICTS IN WEST AFRICA USING ARTIFICIAL INTELLIGENCE TECHNIQUES. Int. J. of Adv. Res. (Apr). 584-596] (ISSN 2320-5407). www.journalijar.com
University of Abomey-Calavi
Benin
