基于多模块模糊贝叶斯网络的恐怖袭击事件量化分析

Quantitative Analysis of Terrorist Attacks Based on Multi-module Fuzzy Bayesian Networks

近年来,全球范围内恐怖袭击事件的发生日益频繁,已经成为许多国家和地区所面临的主要安全问题.首先建立了全球恐怖袭击事件危险性分级指标体系,并基于多模块模糊贝叶斯网络来建立一个基于观测事件'后验概率'的推理模型,从而解决信息不确定时恐怖袭击事件的危险性分级.其次,利用FCM聚类模型对恐怖袭击事件的危险度进行聚类分析,从而依据事件特征对犯罪嫌疑人进行准确锁定.然后,建立了贝叶斯网络恐怖活动预测模型,为全球反恐态势提供有效预测,并利用ArcGIS热点图分析各地区恐怖袭击事件的时空特性.最后,基于VAR模型揭示了不同区域之间在恐怖袭击事件发生时跨区域间的相互影响关系,并提出了相应的反恐建议.

In recent years,terrorist attacks become more and more frequent around the world,which has become a major security problem for many countries.In this paper,a risk classification index system of global terrorist attacks is firstly built.Multi-module fuzzy Bayesian network is used to build an inference model based on the posterior probability of observed events in order to achieve the risk classification of terrorist attacks with uncertain information Secondly,FCM model is used for cluster analysis of the risk degree of terrorist attacks.As a result suspects could be indicated accurately according to the characteristics of terrorist attacks.Then,Bayesian network terrorism prediction model is built to predict the global anti-terrorism situation ArcGIS heat maps are also used to analyses the time-space characteristic of terrorist attacks in different regions.Finally,VAR model is built to explain the interaction relation in different regions when terrorist attacks occurs.Several suggestions of anti-terrorism are also put forward based on the above research.