MDPCA : Moving Dynamic Principal Component Analysis for Non-Stationary Multivariate Time Series

Alshammri, Fayed Awdah M (2020) MDPCA : Moving Dynamic Principal Component Analysis for Non-Stationary Multivariate Time Series. University of Strathclyde, Glasgow.

[img]
Preview
Text (Alshammri-2020-MDPCA-Moving-Dynamic-Principal-Component-Analysis-for-Non-Stationary)
Alshammri_2020_MDPCA_Moving_Dynamic_Principal_Component_Analysis_for_Non_Stationary.pdf
Final Published Version
License: GNU General Public License v2 (Software)

Download (96kB)| Preview

    Abstract

    This function reduce the dimension of non-stationary (and stationary) multivariate time series by performing eigenanalysis on the moving cross-covriance matrix of the extended data matrix up to some specified lag. Notice that thefollowing libraries are needed to be installed before using the MDPCA function: library(roll); library(expm).