一种基于时序数据的动态聚类分析方法
A Method of Dynamic Cluster Analysis Based on Time Series Data
-
摘要: 本文基于灰色关联度和模糊聚类分析原理及多目标决策技术,提出了一种基于时间序列数据的动态聚类分析方法,并以复杂的城市化进程为例加以应用,得出了一些有意义的结果。实践证明,该方法适合于多指标、多时段的(复杂)动态系统建模。Abstract: Based on the principles of grey relative degree,fuzzy cluster analysis and multi-objective decision making technique,the paper presents a method of dynamic cluster analysis through time series data.Taking complicated urbanization process for an example,the method is applied with some significant results.It is proved that the method could be employed to model any complicated dynamic system with multi-factors.