地区涉恐安全风险评估与影响因素研究能够研判地区未来的反恐态势，提高反恐斗争的针对性和资源的管理效率。运用空间统计方法对全球恐怖袭击事件发生的分异特征进行分析，选取恐怖袭击事件数、恐怖袭击死亡人数、恐怖袭击受伤人数和恐怖袭击财产损害程度等指标构建地区涉恐安全风险评估指标体系，提出基于密度敏感稳健模糊核主成分分析的多维面板数据的分级模型。与其他的主成分聚类算法相比较，所提算法的合理性和准确度都有提升。结合主成分方法提取出来的风险水平因子，运用最大似然估计方法建立随机效应面板顺序Logit模型，考察宏观经济和社会发展因素对地区涉恐安全风险的影响，最后对回归模型展开预测绩效评价。 The research on regional terrorism-related safety risk evaluation and its determinants can judge the future situation of regional anti-terrorism, and improve the pertinence of anti-terrorism fight and efficiency of resource management. In this paper, the spatial statistical method is used to analyze the differentiation characteristics of global terrorist attacks. We construct the regional terrorism-related safety risk evaluation index system according to the variables such as number of terrorist attack, number of deaths, number of the injured and extent of property damage in terrorist attacks, and put forward a multi-dimensional panel data classification model based on density sensitive robust fuzzy kernel principal component analysis. Compared with other principal component clustering algorithms, the rationality and accuracy of the proposed algorithm are improved. The influence of macroeconomic and social development factors on regional terrorism-related safety risk is considered by presenting the random effect panel order Logit model based on maximum likelihood estimation method and the risk level factors obtained from principal component method. Finally, the prediction performance evaluation of the regression model is carried out.