基于对象多属性差异的灰色发展聚类方法及应用

Object Multi-attribute Differences Based Grey Dynamic Clustering Method and Its Application

在现实决策问题中,决策对象在不同时期行为状态和所属类型往往呈现一定的发展规律,而现有聚类方法难以充分挖掘聚类对象的发展信息、对象间的关系信息和发展属性的差异信息。为有效处理此类问题,考虑到研究对象的发展趋势、发展行为和发展绝对量与增长量的属性差异,采用GM(1,1)和灰色定权聚类方法,构建了基于对象多属性差异的灰色发展聚类方法,并以我国区域高新技术产业化聚类评估问题为例验证了模型的有效性与合理性。结果表明,所构建模型能够有效描述研究对象呈现发展趋势或未来行为,并实现对研究对象的有效聚类。

The behavior states of decision making objects and their classes from different period often take on a certain development law in the real decision-making problems,however,it is difficult for the existing clustering methods to fully exploit and extract the development information of clustering objects,related information between objects,and difference information of development attributes from clustering objects. To effectively deal with these problems,according to the trends and behaviors of the development and the attribute differences of the development amount and increment amount from the objects,the thought and method of GM( 1,1) and gray fixed weight clustering is utilized to construct a novel object multi-attribute differences based grey dynamic clustering method,and then an example of the clustering problems on the regional high-tech industrialization in China tests and verifies the validity and rationality of the proposed model. The results show that the proposed model can well describe the development trends or future behavior of decision making objects,and realize the effective clustering of the decision making objects.