多目标/多类型恐怖袭击风险评估模型研究  

A Study on Multi-target/Multi-type Terrorism Attack Risk Assessment Model

作　　者：朱炎峰 管涛[1] 卢树华[1] ZHU Yanfeng;GUAN Tao;LU Shuhua(School of Police Information Engineering and Cyber Security,People s Public Security University of China,Beijing 102600,China)

机构地区：[1]中国人民公安大学警务信息工程与网络安全学院,北京102600

出　　处：《中国人民公安大学学报（自然科学版）》2020年第1期64-70,共7页Journal of People’s Public Security University of China(Science and Technology)

基　　金：国家重点研发计划项目(A19808);中央高校基本科研经费项目(2019JKF225)。

摘　　要：全球恐怖袭击频发,对各国公共安全构成了严重威胁,研究多目标/多类型恐怖袭击风险评估,感知和掌握风险态势具有重要意义。基于K-means++聚类分析算法和Python语言编程建立了多目标/多类型恐怖袭击风险函数模型,利用模型对2002~2016年全球恐怖袭击的数据进行聚类分析,将多目标/多类型恐怖袭击风险划分为5个等级,得到针对公民自身和私有财产采用轰炸/爆炸袭击方式的恐怖袭击风险最高。以2017年的数据测试该模型的泛化能力,测试结果准确率达到了94.44%,并与K-means、BP神经网络和决策树等机器学习模型进行了对比分析。结果表明此模型可为恐怖袭击风险评估和防范策略提供一定的参考。The frequent occurrence of global terrorist attacks poses a serious threat to the public security of various countries.It is of great significance to study the multi-target/multi-type terrorist attacks risk assessment and to perceive and master the risk situation.In this paper,a multi-target/multi-type terrorist attack risk function model is established based on the K-means++cluster analysis algorithms and Python programming language.The model is used to perform cluster analysis on GDT data from 2002 to 2016.Divided into 5 risk levels,terrorist attacks that use bombing/explosive attacks against citizens themselves and private property are at the highest risk.Furthermore,the generalization ability of the model was tested by the data of 2017,and the accuracy of the test result reached 94.44%.It was compared and analysed with machine learning models such as K-means,BP neural network and decision tree.The results show that this model can provide a certain reference for terrorist attack risk assessment and prevention strategies.

关 键 词：恐怖袭击 风险评估 K-means++ 

分 类 号：D631[政治法律—政治学]