统计与决策2024，Vol.40Issue(6)：45-49,5.DOI:10.13546/j.cnki.tjyjc.2024.06.008
基于动态综合评价的面板数据聚类方法
Panel Data Clustering Method Based on Dynamic Comprehensive Evaluation
摘要
Abstract
Panel data clustering needs to consider index dimension and time dimension information.This paper attempts to extract the indicator dimension and time dimension information of the panel data from the perspective of dynamic comprehensive evaluation.The dynamic comprehensive evaluation of panel data is carried out by using the vertical and horizontal scatter-degree approach,and the comprehensive evaluation value of index dimension reflecting the level of samples is obtained.Then,the dynam-ic comprehensive evaluation model with velocity characteristics is used to evaluate the incremental velocity of the sample,and the comprehensive evaluation value of time dimension reflecting the fluctuation size and changing trend of the sample is obtained.Fi-nally,the comprehensive evaluation value of the two dimensions is systematically clustered.The result of example analysis shows that the proposed method has better clustering effect.关键词
面板数据/动态综合评价/指标维度/时间维度/聚类分析Key words
panel data/dynamic comprehensive evaluation/index dimension/time dimension/cluster analysis分类
数理科学引用本文复制引用
梁锦玲,邓光明..基于动态综合评价的面板数据聚类方法[J].统计与决策,2024,40(6):45-49,5.基金项目
国家自然科学基金资助项目(71963008) （71963008）
