摘要
本文选取四川省某县(区)公安局基于"一标三实"对当地18个镇部分项目进行的考核得分,运用SPSS软件的统计分析和建模方法对数据进行分析,通过聚类分析将该区划分分为四类.并利用多元线性回归方法将最后的均值得分作为因变量,以标准地址、实有单位、实有人口、质量考核作为自变量,进行回归分析.通过两次对模型的修正,最后得出了一个较优模型.
This paper selects public security burea data of examination score of 18 local townships part of examination scores( based on "Standarized Address and Three Facts"). SPSS is used for data processing and modeling. On the basis of cluster analysis,this region is analyzed into four categories. Then we define the final average score as the dependent variable,standard address,actual unit,actual population,quality inspection as the independent variable. Finally we conduct a model via multiple linear regression method and correct it twice before it reaches a relatively optimal model:Mean score = e^8. 113 +0. 142* Ln( population) +0. 253* Ln( workplace) +1. 171*( Ln( quality))^ 2-4. 64* Ln( quality)
出处
《洛阳师范学院学报》
2015年第8期11-14,共4页
Journal of Luoyang Normal University
关键词
一标三实
聚类分析
多元线性回归
"standardized address and three facts"
cluster analysis
multiple linear regression