This study aims to identify the determinants of terrorism, namely the number of attacks, and generate an explanatory model for the occurrence of terrorist acts, having been considered in the analysis economic, socio-demographic, political, and religious variables, 173 countries and a period of 20 years. Through the application of decision trees, namely the CART algorithm, several variables of the different dimensions have importance for the explanation of the number of attacks, being the socio-demographic that better explain this terrorism metric. The regression tree model does not show a very high explanatory capacity, but given the complexity and subjectivity involved in the phenomenon of terrorism, we can consider that the estimated model is a further contribution to the literature and to the understanding of the terrorism drivers.

--