An Improved Interval-type Symbol Data Principal Component Analysis

DOI

10.2991/smont-19.2019.45How to use a DOI?

Keywords

crincipal component analysis; interval number; correlation matrix

Abstract

For interval sample data, a new principal component analysis method based on empirical correlation matrix is proposed. The method is to treat the interval samples as binary variables obeying the normal distribution, and obtain the empirical description statistic[1,2]—experience covariance matrix[3] (empirical correlation matrix) through the joint density distribution, and then calculate eigenvalues of empirical covariance matrices (experience correlation matrices) and orthogonalized unit eigenvectors[4]. And then, through a certain operation, the interval unit eigenvector is obtained. Further obtaining the main components. Finally, a concrete example is given to verify it. This method discards the assumption of traditional uniform distribution, and regards interval samples as binary variables obeying normal distribution, which is more universal and more accurate.

Copyright

© 2019, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Weiguo Zhang
AU  - Qingxian Liu
PY  - 2019/04
DA  - 2019/04
TI  - An Improved Interval-type Symbol Data Principal Component Analysis
BT  - Proceedings of the 2019 International Conference on Modeling, Simulation, Optimization and Numerical Techniques (SMONT 2019)
PB  - Atlantis Press
SP  - 202
EP  - 206
SN  - 1951-6851
UR  - https://doi.org/10.2991/smont-19.2019.45
DO  - 10.2991/smont-19.2019.45
ID  - Zhang2019/04
ER  -