首页 | 本学科首页   官方微博 | 高级检索  
     检索      

卢龙县低山丘陵区灌丛的聚类分析
引用本文:王德艺,陈连涛.卢龙县低山丘陵区灌丛的聚类分析[J].河北林果研究,1990(1).
作者姓名:王德艺  陈连涛
作者单位:河北林学院林学系 (王德艺),卢龙县林业局(陈连涛)
摘    要:本文应用聚类分析的方法对卢龙县低山丘陵区112个落叶阔叶灌丛样地资料进行了分类。样地间的相异系数采用欧氏距离、聚合策略选用最近邻体法、最远邻体法、中线法、形心法和平方和增量法。结果表明,最近邻体法空间压缩性较强,中线法和形心法存在逆转的缺点,最远邻体法和平方和增量法对样地资料的利用率最高,错分类最少,是较为理想的分类方法。文中还按平方和增量法将112个灌丛样地资料分为9个群系,23个群丛组。

关 键 词:聚类分析  灌丛

THE CLUSTER ANALYSIS OF DECIDUOUS BROAD-LEAVED BRUSH-WOOD IN THE HILLY DISTRICT IN LuLONG COUNTY
Wang Deyi.THE CLUSTER ANALYSIS OF DECIDUOUS BROAD-LEAVED BRUSH-WOOD IN THE HILLY DISTRICT IN LuLONG COUNTY[J].Hebei Journal of Forestry and Orchard Research,1990(1).
Authors:Wang Deyi
Institution:Wang Deyi (Hebei Forestry College) Chen Liantao (Forestry Bureau of Lulomg County)
Abstract:This paper classifies 112 releves ( 24 species ) of the deciduous broad-leaved brush-wood in the hilly district in Lulong county using the cluster analysis method. The distance formula Euclidean is adopted to calculate similarity between the releves. "We use five methods in the cluster analysis. They are, nearest neighbour, furthest neighbour,median, centroid and incremental sum of squares m'ethod. The results show that the nearest neighbour method has weak point in space-contracting, the median and centroid methods have the defect of reverse, the furthest neighbour and incremental sum of squares are the better classification nrethods, for they have the highest utilization ratio of releves and the least erroneous classification rate. Based on the result of incremental sum of squares method, 112 releves are divided into 9 formation, 23 association group.
Keywords:cluster analysis  brush-wood  
本文献已被 CNKI 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号