A PCA for interval-valued data based on midpoints and radii

被引：53

作者：

Palumbo, F ^{[1
]}

Lauro, CN ^{[1
]}

机构：

[1] Univ Macerata, Dip Ist Econ & Finanziarie, I-62100 Macerata, Italy

来源：

NEW DEVELOPMENTS IN PSYCHOMETRICS | 2003年

关键词：

interval data; principal components; graphical representations;

D O I：

10.1007/978-4-431-66996-8_74

中图分类号：

O1 [数学];

学科分类号：

0701 ; 070101 ;

摘要：

In this paper, we propose a new approach to Principal Component Analysis, for interval-valued data. On the basis of the interval arithmetic we show that any continuous interval can be expressed in terms of a midpoint (location) and of a radius (variation). Moving from this result, we propose a well suited factorial analysis, which exploits this characteristic of interval data. Both the location and variation information are represented on maps.

引用

页码：641 / 648

页数：8