为得到尽可能多的系统信息,通常是选择尽可能多的与系统有关的指标来描述系统,但大量的统计指标会带来统计、分析的难度,由于系统的元素在系统中的地位和作用是不平衡的,若通过一些方法找到系统中起决定作用的指标就可以简化问题。利用系统核与核度理论可以对大量指标进行筛选,找出与所研究问题相关的关键指标,但对于如何将一些系统转化为图以及系统的核不唯一性问题造成了系统核理论难以在实践中得到应用,本文利用相关分析给出了将实际系统转化为图的方法,结合关联聚类分析解决了系统核理论中的多核问题,提出了系统核的选择模型,利用模型得到了武汉市房地产系统的核元素,将这些核元素编制为扩散指数对武汉市房地产1990-2001年期间的发展规律进行分析,结果表明,得到的扩散指数曲线与武汉市房地产在1990-2001年的发展变化是相一致的,因而利用模型得到的核元素能够很好地描述武汉市房地产系统。
Due to the lack of uniqueness of the system core, the theory of system core has a problem with the core-choosing in its application. This paper gives a method to turn a system into a graph and formulates a model to select system core within the cluster analysis and grey correlative analysis. Taking the real estate system of Wuhan city as an example, this paper sets up the diffusion index with the elements of system core by the model, and the diffusion index curve indicates that the real estate arise since 1991 to the peak in 1993 and then decline to the bottom in 1996, then fluctuated in a small range, but it has become smooth since 1999, however, there is a slight upward trend since 2000. This conclusion demonstrates the validity of the model.