基于Floyd改进算法的北京产业结构网络强关联模糊聚类分析

Fuzzy Cluster Analysis of Strong Association in Beijing Industrial Structure Networks based on Improved Floyd Algorithm

根据复杂网络建模理论,采用投入产出直接消耗系数构建产业结构网络模型,并改良传统的Floyd算法构造强关联矩阵,使该矩阵具有收敛性、唯一性、最优化和无标度等特点,从经济技术信息传递快速性角度出发,描述产业部门之间的技术经济关系,最后运用北京市连续8年的4组投入产出数据进行产业部门强关联模糊聚类分析。

In this paper input-output basic matrix of Beijing was made use of to build a sort of directed and weighted network named industrial structure network based on the theory of complex networks. Then, an improved Floyd algorithm was chosen to form a kind of characteristic matrix with features of Convergence, Uniqueness, optimization and scale-free, which reflected strong association among industrial sectors and aimed at researching techno-economic relations in terms of information transmission. Finally, model and formula were used to do fuzzy cluster analysis with 4 groups of data from 2002 to 2010 of Beijing.