基于MLE的恐怖组织袭击行为模式实证研究  

Empirical Studies on Attack Patterns of Terrorism Groups Based on MLE

作　　者：赵法栋[1] 庄弘炜[1] 金振兴 

机构地区：[1]武警工程大学装备工程学院,西安710086 [2]武警工程大学理学院,西安710086

出　　处：《复杂系统与复杂性科学》2014年第4期19-22,40,共5页Complex Systems and Complexity Science

基　　金：国家社会科学基金(11BZX016)

摘　　要：为揭示恐怖组织的袭击行为模式，对袭击数量超过500件的恐怖组织袭击行为进行了统计实证。在袭击时间间隔服从幂律分布的假设下，采用最大似然估计法对其幂律指数进行估计，并采用Kolmogorov—Smirnov检验法进行真伪性检验。结果表明：各组织发动袭击的时间间隔均服从幂指数为1．8～3．1的幂律分布；恐怖袭击的阵发性与恐怖组织创立的早晚、生命周期的长短无关；组织平均活跃性与幂律指数并不具有正相关关系，组织规模的大小是影响其幂律指数的重要因素。To reveal the attack pattern, terrorism groups which attacked more than 500 events were empirically studied. Under the hypothesis that their interevent time follow power-law distri- bution, the maximum-likelihood estimation was applied to estimate the power-law exponent, and then the Kolmogorov-Smirnov（KS） statistic was used to test the hypothesis. The results show that the interevent time distributions of all the groups are power-law; The burst of terrorism attacks has no relationship with the foundation time and the lifetime; There is no positive correlation between average activity and power-law exponent and the group size is the main factor to affect it.

关 键 词：恐怖组织 幂律分布 平均活跃性 实证研究 最大似然估计 

分 类 号：C812[社会学—统计学]