チャーノフの顔形グラフの応用研究(I)

概要

論文の詳細を見る
The purpose of this study is to investigate the method for assigning multivariate data to the features in the Chernoff's face graph. The authors present the method in which the variables are analyzed by means of factor analysis and, in order of magnitude of the absolute value of factor loading of common factor are assigned to the features which has the large contribution rate to visual reaction. The following results are obtained through an application to managerial index data of Japanese banks. (1) Observers could objectively and easily judge the state of a common factor by observing the face graph if the meaning of the common factor is explained. (2) It is generally considered that an individual or a sample is in the same state if its factor score is the same, but it is sometimes observed that the expressions on the face graph vary, even if they are of the same factor score. As a result, Chernoff's face graph is available as an auxiliary method of exploratory data analysis.
明治大学の論文