灾害学2016，Vol.31Issue(3)：184-189,6.DOI:10.3969/j.issn.1000-811X.2016.03.031
基于贝叶斯网络的恐怖袭击预警模型研究
Research on Terrorist Attack Warning Model Based on Bayesian Network
摘要
Abstract
Terrorism early warning can be more effective prevention and control of terrorist attacks has brought worldwide attention.However,the terrorist attacks in our country establish a direct warning model,facing the sam-ple data is insufficient,so that the model cannot be accurately established.Therefore,we propose a method to solve this problem by using the Bayesian network with the principle of case adaptation.In order to guarantee its case adaptation,we specify the rules of data selection and node selection,and establish a Bayesian network model based on foreign terrorist attack samples,and is tested for its practicality and versatility.Then according to the principle of case adaptation,combined with the actual data,we use the EMalgorithm to update the parameters of learning,and then modify the model.After verification,the results show that the model accuracy is greatly im-proved,which is more in line with the actual situation of our country.Finally,the model warning a terrorist attack, casualties highlighting conclusions are drawn,and the results provide effective warning information for the decision makers.关键词
恐怖袭击/预警/贝叶斯网络/案例适配Key words
terrorist attack/warning/Bayesian networks/case adaptation分类
资源环境引用本文复制引用
傅子洋,徐荣贞,刘文强..基于贝叶斯网络的恐怖袭击预警模型研究[J].灾害学,2016,31(3):184-189,6.基金项目
天津市哲学社会科学规划资助项目 （）
