摘要
聚类分析是统计学中建立分类的多元统计分析方法,它能把一批样本按照诸多特征,依据性质上的亲疏在没有先验知识前提下自动分类,产生若干分类结果。采用聚类分析的办法对山东省各个城市火灾统计数据进行研究,用以确定各个城市火灾特征的显著性差异,然后分类别挖掘火灾的制约因素,既可以使我们制定消防政策方针时具有针对性,同时计算结果还可以评价地区消防工作和消防政策的有效性.
Clustering analysis is a multivariate statistical classification statistical analysis method ,it can take a number of samples in accordance with many features ,based on the nature of the relationship in the absence of prior knowledge under the premise of automatic classification ,some classification re‐sults .From the overall perspective ,the analysis time for each province city fire situation ,can use clus‐tering analysis to study the province each city fire statistical data ,to determine the significant difference between each city fire characteristics ,and then classified mining factors of fire ,we can have in the development of fire protection policy policy ,effectiveness and result can also fire prevention and fire area evaluation policy .
出处
《聊城大学学报（自然科学版）》
2016年第3期91-93,共3页
Journal of Liaocheng University:Natural Science Edition
基金
国家自然科学基金项目(51075244)资助
关键词
层次聚类
城市火灾形势
分析评价
hierarchical clustering
city fire situation
analysis and evaluation
