Research On The Prediction Model Of Regional Terrorism Situation

With the continuous spread of terrorism,national security is facing major challenges,and counter-terrorism emergency management needs to break through.At present,terrorism has become an important security risk threatening the economic development,social stability,people’s life and property security of all countries,which has a farreaching impact on the integration of global political and economy.In particular,taking the 9/11 incident as a turning point,terrorism is characterized by transnational,networked,multi-attribute,etc.According to the characteristics of terrorist activities,using appropriate data analysis and mathematical model,mining the occurrence and development of terrorism,and on this basis,warning risk is helpful to prevent terrorist events to a certain extent.At present,most of the modeling perspectives of anti-terrorism research focus on the prediction of the occurrence of terrorist events.It is still to be explored for the prediction of the regional overall terrorism situation.This study proposes the conceptual framework of terrorism situational awareness based on the basic idea of situational awareness,describes the relevant concepts and process architecture,and provides theoretical sources for the top-level design and research of terrorism situation research.On the basis of this research,this paper carries out the research on the method of terrorism situation prediction.First of all,this paper introduces Mapping Knowledge Domains to construct the terminology co-occurrence network to identify the key situation elements,and transforms them into quantitative indicators.The Euclid algorithm is introduced to modify the grey correlation method,and the indicators is selected to build the index system of terrorism situation prediction,which lays the foundation for the establishment of situation prediction model.Then,the whale optimization algorithm is introduced to improve the least squares support vector regression machine,and the terrorism situation prediction model based on WOA-LSSVR is established.We use the whale algorithm’s excellent global optimization ability to optimize the parameters of support vector regression,and use the optimized model to predict the terrorism situation of various countries.Finally,taking the countries along the "belt and road" as the object of empirical analysis,the 2008-2018 years data are used to predict the terrorism situation index.The results are compared with other models to verify their performance.The conclusion shows that this modeling method has stronger fitting degree and higher prediction accuracy,and can solve the problem of terrorism situation prediction excellently.To some extent,this study can assist the anti-terrorism action department to make scientific prevention strategies,and help decision makers to improve the efficiency and accuracy of decision-making.