The Establishment of Green Industry Evaluation Index System Based on Dynamic Clustering - an Empirical Study from Dalian in China

Use the method of spearman correlation analysis and dynamic clustering to delete indicators and build an evaluation index system of the green industry. The specialties and contributions of this paper lie in three aspects. Firstly, screen indicators according to the different effects on the clustering results of the evaluation objects from having or not having specific indicators, and delete the indicators that have no influence on the evaluation objects clustering results so as to guarantee that every selected indicator has effects on the clustering results. Secondly, delete the indicators with high correlation coefficient in the same criteria by spearman correlation analysis to avoid information duplication of the indicators. Thirdly, the eventually established indicator system reflects 94.18% of the original information by 27% of the indicators. In the future, we can establish a green industry evaluation software system. By using this system, the government can evaluate the development of green industry easily and establish policies to balance the development of different area.