论文部分内容阅读
古生物地理的划分,通常有综合比较法和数字统计法两种.前者是传统的方法,而数字统计法是常用的二态变量聚类分析法:以地层为样本,地层中包含的种属为变量,作Q型聚类分析.我们通过若干资料的计算,取得了较为满意的结果.但是笔者在计算过程中发现,当形成大类时,相关系数往往降到很低.作为数字分类方法的探索,对同样的资料我们采用模糊聚类法分类,结果基本上与Q型聚类法一致,而且在某些地区与综合比较法的结果更吻合.还有一个明显的特点是:在和Q型聚类法分成同样大类的情况下,模糊聚类法给出了较大的亲近度.
Paleogeography division, usually there are two kinds of comprehensive comparison method and digital statistics.The former is a traditional method, and digital statistics method is commonly used two-state variable cluster analysis method: the formation as a sample, the strata contained in the species is Variables, for Q-type cluster analysis.We obtained some satisfactory results through the calculation of a number of data.However, I found in the calculation process, when the formation of large categories, the correlation coefficient tends to be low.As a method of digital classification Exploration, the same information we use fuzzy clustering method, the results are basically consistent with the Q-type clustering method, but also in some areas with the results of the comprehensive comparison method is more consistent.Another obvious characteristic is: in and Q Type clustering method into the same category, the fuzzy clustering method gives a larger degree of closeness.