Spatial Correlation Research Based On Agglomeration Effect Recognition And Gi* Constraints

Spatial Dubin Model(SDM)is a econometric model for studying the interaction or association of elements from the perspective of geographical location and spatial relations.Based on the clustering measure,this paper proposes a clustering effect recognition method based on "two minimum" methods,which solves the problem of difficult to identify the agglomeration area and thus detects the difference of agglomeration.At the same time,on the basis of SDM application,the method of constructing spatial weight matrix(SWM)under Gi * statistic constraint is proposed,which solves the problem of fuzzy neighborhood of neighborhoods and explores the reasons of agglomeration differences.And,this paper studies the spatial distribution pattern of productive service enterprises by using the method proposed in this paper.The main research contents are as follows:(1)The method of cluster effect identification based on minimum spanning minimum convex hullIn view of the problem of parameter dependence of spatial clustering measures,this paper proposes a clustering effect recognition method based on MST-MCH algorithm.Based on the spatial object data structure and the improved method of cluster analysis,the technology route of sub tree construction,subtree separation,convex hull formation and cluster identification is formed.The experimental results show that the method can identify and classify the gathering area very well,and the classification is performed,and the results are more reliable.(2)The study of the spatial dubin model under the constraint of Gi* statisticIn view of the randomness of the neighborhood selection in the SDM model construction,this paper proposes a neighborhood selection method under the constraint of Gi* local statistics.On the basis of simple weight and distance weight method,this paper introduces the spatial statistics,and draws on the idea of exponential maximization proposed by Anselin,adopts the strategy of maximizing the Gi* statistic,and solves the problem of difficult selection of related neighborhood The improvement of SWM construction method lays the foundation for the construction of SDM and promotes the development of econometrics in theory.(3)The study of the spatial pattern of producer service enterpriseExperimental study for the Beijing Tianjin Hebei production service enterprises,collect data,using industrial geographic concentration of population geography extended concentration,Gini coefficient,location quotient and MST MCH algorithm by analyzing the differences,constructing SWM using Gi* statistic correlation neighborhood constraints,so as to study the producer service enterprise SDM,the effectiveness and superiority of the proposed method is a test;on the other hand,clear the cause factor or index difference,put forward some suggestion,provides the reference for the future development.