首页 | 本学科首页   官方微博 | 高级检索  
     

Bayes判别在巨猿分类中的应用
引用本文:王令红,方开泰,郑玉颖. Bayes判别在巨猿分类中的应用[J]. 古脊椎动物学报, 1981, 0(3)
作者姓名:王令红  方开泰  郑玉颖
作者单位:中国科学院古脊椎动物与古人类研究所,中国科学院应用数学研究所,中国科学院应用数学研究所
摘    要:<正> Bayes判别分析是多元分析中的方法之一。多元分析是用数理统计的理论研究多项指标问题的一种方法,它的理论在本世纪30年代就得到相当完整的发展。最早用于考古学和古生物学的研究之中,后来应用面越来越广,目前已广泛地应用于地质、医学、心理学、行为科学、标准化、经济、海洋、农业等方面,并取得了很好的效果。通过电子计算机的运算,采用数学方法进行分类,已成为当代生物分类的三个主要流派(综合学派、分支学派和表现型学派)之一,数量分类即是表现型学派据以分类的实质内容。就世界范围而言,数量分类方法已经进入了生物各大门类的分类之中,甚至于渗透到生物学的其他领域。多元分析是数量分类中目前比较理想的方法。


AN APPLICATION OF BAYES DISCRIMINANT ANALYSIS IN DETERMINING THE SYSTEMATIC POSITION OF GIGANTOPITHECUS
Abstract:The systematic position of Gigantopithecus in Hominoidea has been a subject of heated discussion among anthropologists for a long time. Several foreign scholars have discussed this problem from the view-point of numerical taxonomy. We try to reexamine the status of Gigantopithecus by means of multivariate discriminant analysis. In this paper we introduce the method of Bayes discriminant analysis and show the result we have obtained through its application.Gigantopithecus was compared to both modern man and the gorilla to determine whether it resembles the former or the latter. Our discriminant analysis is composed of two parts: (1) mandibles (5 variables: bicanine breadth, symphyseal height, corpus height and breadth at M_2, P_2—M_2 chord) ; (2) teeth (12 variables: length and breadth of I_2 through M_2). Male and female specimens are dealt with respectively.The result of discriminant analysis shows that the mandibles of both male and female Gigantopithecus are similar to those of gorilla, but the teeth, in contrast, more closely resemble those of modern man.The resemblance between Gigantopithecus and modern man may be regarded as an affinity, rather than a convergence. So Gigantopithecus should have ancestral status as a creature translated from ape to man, which was no longer arboreal, but lived on the ground. The change of food source led to the change of the dental morphology and its size firstly and directly.
Keywords:
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