两维图论聚类法在农业区划中的应用——以山东省十七地市为例  

Application of Two Dimensional Graph Theory-clustering Method in Agricultural Regionalization——Take 17 places cities in Shandong Province as a case

作　　者：谷晓岩[1] 李凤英[2] 张燕[1] 

机构地区：[1]鲁东大学地理与规划学院,山东烟台264025 [2]江西农业大学国土资源与环境学院,江西南昌330045

出　　处：《安徽农学通报》2009年第3期62-64,59,共4页Anhui Agricultural Science Bulletin

摘　　要：聚类分析是数理统计中研究"物以类聚"的一种方法,它将一批样品或变量按照它们在性质上的亲疏程度进行分类。农业区划的评价从结果上看其实也是一种分类过程。农业区划不但要考虑评价单元的自然、经济和社会效益,而且要保持空间上的连通性和行政区界的完整性。本文利用两维图论聚类法对农业区划进行分析,确定最佳的区位优势、分类及组合,以便于农业的集约化和耕作的合理化,较好的配置农业的产业结构,为农业可持续发展和实现合理的农业地域分工提供科学依据。Clustering analysis is a method in the mathematical statistic studying ＂the birds of a feather flock together＂. It car- ries on one batch of samples or the variable to classify according to their friend or stranger degrees in character. Looked from the results, the appraisal of the agricultural regionalization is also one kind of classification process. The agricultural region- alization not only must consider natural, economic and social benefit of the appraisal unit, moreover must maintain the con- nectivity on the space and the integrality of the administrative regional boundaries. In this paper, the agricultural region is de- limited by using the two dimensional graph theory -clustering method. The optional location superiority, the classification and the combination were determined in order to make the agricultural intensification and cultivation rationalization and good disposition industrial structure, which will provide the science basis for agricultural sustainable development and realize the reasonable territorial division of labor in agricuhure.

关 键 词：两维图论聚类法 农业区划 山东省 

分 类 号：S11[农业科学—农业基础科学]