Graphical Representation of Multivariate Data by Means of Asymmetrical Faces
在Chernoff脸谱图基础上，提出左右脸参数可独立变化的新脸型，能表示36个变量，用于多元配对比较和异常值视觉搜索，比原版更真实、退化少且参数间干扰小。
Abstract Proceeding from the idea of Chernoff (1971, 1973) of representing multivariate data by faces, a new face is proposed in which the face parameters of the left and the right side can be varied separately. The new face can be applied in the same way as the usual Chernoff face (with 36 instead of 18 representable variables). Special applications of the new face are the representation of multivariate paired comparison and the visual search for outliers. Several examples illustrate these applications. Compared with Chernoff's face, the new construction is more realistic, degenerates less, and avoids as much as possible mutual influences of the different face parameters.