Skip to Main Content
Recently, terrorist or rebel activity is experienced in many parts of the world. The objectives of the research are to design and develop an accurate predicting the distribution range radius and elapsing time of a terrorist situation. An Analytical Network Process (ANP) is used to classify salient quantitative and qualitative factors of the unsettled area, or terrorist behaviour. Then, the resulting Artificial Neural Networks (ANNs) are used to set up initial factor weights. The ANNs model is trained and tested for verification and validation against a historical data set. Improvised Explosive Device (IED) events from 2007 to 2011 in the capital district of Yala province, the southern part of Thailand, are used for testing the experiment. The proposed decision support methodology emerges as capable of predicting the distribution range radius and the elapsing time from the previous incident. The ANP technique can analyse and weight the complex quantitative and qualitative criteria to yield basic inputs to the specified ANNs. With initial set up architecture by the ANP weighting results, the contribution of this research lies in the proposed ANNs' capacity to predict a terrorist incident. Further, assessment of the Explosive Ordnance Disposal Mobile Unit (EODMU) is simulated by using the prediction results from the case study. As a result, during 2007 to 2012, with more opportunity to detect and defeat the next IED terrorist incident, the radius risk operation area of the capital district of Yala province can be reduced from some 8.9 km. to 5 km. By applying the model, the Thai Government can focus on a smaller area, resulting in reduced expenditure of military assets with no cost in additional casualties.