基于k-means聚类与粗糙集算法的指标筛选方法研究  

Indicators Screening Method Based on k-means Clustering and Rough Set Algorithm

作　　者：张立军[1] 高春晓 ZHANG Li-jun;GAO Chun-xiao(College of Finance and Statistics, Hunan University, Changsha 410079, China)

机构地区：[1]湖南大学金融与统计学院,湖南长沙410079

出　　处：《运筹与管理》2020年第12期8-12,共5页Operations Research and Management Science

基　　金：国家社会科学基金资助项目(14BTJ003)。

摘　　要：针对多属性决策中指标的信息重复和不确定性问题,提出了一种基于改进的k-means聚类与粗糙集算法相结合的指标筛选方法。首先,定义样本的空间分布密度,实现初始聚类中心优化的k-means算法,对连续型指标进行离散化处理;然后利用粗糙集的相对约简原理进行指标约简,删除存在信息重复的冗余指标,并结合绿色经济指标体系构建的案例验证了该方法的合理性和有效性。Aiming at the information duplication and uncertainty of indicators in multi-attribute decision,an indicators screening method is proposed based on improved k-means clustering and rough set algorithm.Firstly,we define the spatial distribution density of the sample,and implement the k-means algorithm of the initial cluster center optimization,so the continuous indicators are discretized.Then we use the relative reduction principle to reduce the indicators,and delete the redundant indicator with repeated information.Combined with the case of green economic indicator system,the rationality and effectiveness of the method are verified.

关 键 词：K-MEANS聚类 粗糙集 指标筛选 

分 类 号：F205[经济管理—国民经济]