Comparison of Machine Learning Approaches in the Prediction of Terrorist Attacks

Global terrorism means the use of intentionally indiscriminate and illegal force and violence for creating terror among masses in order to acquire some political, monetary, religious or legal goals. Identification of these ideologies and prediction of future attacks has proven to be of the greatest importance but is time-consuming. This paper focuses on analyzing the historical dataset of Global Terrorism Database and predicting the factors that might give blow to an increase of terrorism. Various Data mining techniques and machine learning algorithms like Support Vector Machine, Random Forest, and Logistic Regression etc. have been used to analyze the dataset and carry out predictions like the success of a particular attack, predict the group that carried out an attack and effect of the external factors. A detailed comparison for each algorithm is carried out to attain the most significant results.