对恐怖袭击事件记录数据的量化分析与研究
    点此下载全文
引用本文:王向爱1,庄元强 2,谢为顿3,周金华4,王利平4.对恐怖袭击事件记录数据的量化分析与研究[J].经济数学,2019,(3):95-103
摘要点击次数: 71
全文下载次数: 0
作者单位
王向爱1,庄元强 2,谢为顿3,周金华4,王利平4 (1.湖南大学 工商管理学院湖南 长沙 4100822.湖南大学 机械与运载工程学院湖南 长沙 4100823.湖南大学 信息科学与工程学院湖南 长沙 4100824.湖南大学 数学与计量经济学院湖南 长沙 410082) 
中文摘要:恐怖主义是人类的共同威胁,利用数据挖掘可以为反恐防恐提供有价值的信息支持.基于数据挖掘的思路,从恐怖袭击事件中提取能描述危险程度的特征属性,构建量化分级模型,并考虑准确率评价指标进行优化.通过组内平方和法分析改进高斯混合模型(GMM),对恐怖组织进行聚类分析,侦查出潜在最相关的嫌疑人.建立相关模型结合统计分析,得到恐怖袭击发生的主要原因、时空特性和蔓延特性,并对未来全球反恐态势进行预测,帮助反恐组织提高反恐的精准性和打击能力.
中文关键词:应用统计数学  恐怖袭击事件  数据挖掘  GMM聚类分析
 
Quantitative Analysis and Research on the Recorded Data of Terrorist Attacks
Abstract:Terrorism is a common threat to mankind. The purpose of this paper is using data mining to provide valuable information support for counter-terrorism and terrorism prevention. Based on the idea of data mining, the feature attributes that can describe the degree of danger are extracted from the terrorist attacks, and the quantitative grading model is constructed, and the accuracy rate evaluation index is considered to optimize. Then, the clustering method was optimized by the within-group square method, and the Gaussian mixed model (GMM) model is used to cluster the terrorist organizations to detect the suspects which is the most relevant. In addition, through statistical analysis and established mathematical models, the main causes, spatio-temporal characteristics and spread characteristics of terrorist attacks are obtained, and the future global counter-terrorism situation is predicted to help anti-terrorism organizations improve the accuracy and strike ability of anti-terrorism.
keywords:Sapplied statistical mathematics  Terrorist attack  Data mining  GMM clustering analysis
查看全文   查看/发表评论   下载pdf阅读器