地理科学 ›› 2001, Vol. 21 ›› Issue (6): 525-530.doi: 10.13249/j.cnki.sgs.2001.06.525

• 论文 • 上一篇    下一篇

地域分异的灰色聚类方法研究——以温州市为例

黄劲松, 周生路, 彭补拙   

  1. 南京大学城市与资源科学系, 江苏 南京 210093
  • 收稿日期:2000-05-09 修回日期:2000-09-04 出版日期:2001-11-20 发布日期:2001-11-20
  • 基金资助:
    国家自然科学基金资助项目(49801008)。

A Discussion on Method of Grey Incidence Cluster Applied to Geographic Zone Differentiation ——A Case of Wenzhou, Zhejiang Province

HUANG Jin-song, ZHOU Sheng-lu, PENG Bu-zhuo   

  1. Department of Urban and Resources Sciences, Nanjing University, Nanjing, Jiangsu 210093
  • Received:2000-05-09 Revised:2000-09-04 Online:2001-11-20 Published:2001-11-20

摘要: 直接采用灰色关联聚类进行地域分异,效果并不理想,如果对灰色关联聚类算法进行改进,则可取得满意的结果。结合温州市实际情况,提出地域分异指标的确定原则,采用改进后的灰色关联算法,得出温州市所辖11个区、县、市的聚类谱系图,并将温州市所辖11个区、县、市分为沿海平原城镇-工业发达区、沿海平原农耕区、西部山地生态脆弱区和东部海岛区,对每区的景观特征进行了描述。

Abstract: Geographic Zone Differentiation (GZD) is one of the basic principles of geography research. Grey cluster method is useful and effective in research of classification. But if the Method of Grey Incidence Cluster (MGIC) is applied to research of GZD directly, the result is not completely correct and is not in accordance with the fact. The reason is that the definition of Absolute Degree of Incidence (ADI) is not suitable to GZD. When it is defined as follows, the conclusion drawn by applying MGIC to GZD is correct and satisfying. In these formulas, εij is the ADI between sample i and sample j; xi(k) is the value of the index k of sample i. How to choose and decide the indexes of GZD is also brought out, whose principles are: 1) the index is dominant and effective; 2) the values of the index in varied samples are disparate; 3) the indexes are independent and can't be relevant in linearity; 4) the indexes are distinctly different in reflecting geographic sight.

中图分类号: 

  • P94/X823