多个相异性矩阵的数据分析(英文)

Data Analysis for Several Dissimilarity Matrices

本文给出了分析多个相异性矩阵的三种方法.首先找到了一种图表示,使我们对所有相异性矩阵有一个总体的了解;其次定义了一个新的相异性矩阵,它可以看作是对所有原始相异性矩阵的一个折衷处理;最后提出了一种MIMU方法.在文中我们还对由上述方法得到的坐标图进行了比较.

In this paper, we give three methods to analyse several dissimilarity matrices which indicate the dissimilarities between stimuli from distinct sets in different individuals. We find a graphical representation which allows us to do an overall comparison of the dissimilarity matrices. A new dissimilarity matrix is defined, which can be considered as a goodcompromise between all the initial dissimilarity matrices. MIMU method is presented to deal with a set of dissimilarity matrices. The configurations calculated by above methods are compared mutually.