A classification model for terror incidents by affiliation category of perpetrators

A terror incident could be perpetrated by either a lone wolf who acts on their own or affiliated terrorists who work for a terror group. In this study, the data from the global terrorism database and the artificial neural network algorithm were employed to construct a classification model that could predict the probable affiliation category of the perpetrator(s) of a terror incident. The model uses information such as type of attack, casualty figure, claim of responsibility, and damage to property to distinguish a lone wolf attack from a terror group attack. Various metrics of model diagnostics were employed to test the suitability of the model for predictions, and it yielded a balanced classification accuracy of 85%. The model adds another dimension to the existing criteria for terrorism classification. Further, the model could serve as a useful tool in the study of terrorism and counterterrorism.