论文部分内容阅读
一、引言聚类分析是一种应用十分广泛的多元统计方法。在地质学领域应用尤为普遍,几乎国内开展的各项矿床统计预测工作,都进行了某种聚类分析。其目的是根据地质体的多种地质或地球化学标志的相似性,按其某种数量指标的亲疏程度,逐步聚合而进行分类。对于Q型(关于样品的)聚类而言,在样品较少的情况下,其方法简便,结果直观;但在样品数巨大的情况下,谱系图结果的分析解释,仍是一个很复杂的问题。谱系图的判读,其结果往往与专业人
I. INTRODUCTION Clustering analysis is a widely used multivariate statistical method. In the field of geology is particularly common in the field of application, almost all the statistics carried out in the country to predict the work, have carried out some cluster analysis. Its purpose is to classify according to the similarities of various geologic or geochemical signs of geologic bodies according to the degree of relative density of some quantitative indicators. For type Q (for sample) clustering, the method is simple and the result is intuitive with fewer samples; however, the analysis of the results of the phylogenetic tree is still a complex one with a large number of samples problem. Pedigree diagram interpretation, the results often with professionals