Open AccessThe PCA-seq method applied to analyze of the dynamics of COVID-19 epidemic indicators
Author(s) -
В. М. Ефимов,
D A Polunin,
V. Yu. Kovaleva,
Kirill Efimov
Publication year - 2021
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1715/1/012025
Subject(s) - principal component analysis , univariate , multivariate statistics , covid-19 , series (stratigraphy) , euclidean distance , trajectory , time series , computer science , matrix (chemical analysis) , pattern recognition (psychology) , mathematics , data mining , statistics , artificial intelligence , medicine , paleontology , physics , materials science , disease , pathology , astronomy , infectious disease (medical specialty) , composite material , biology
In time series analysis using the SSA method, a univariate series is converted into the multivariate one by shifts. The resulting trajectory matrix is subjected to principal component analysis (PCA). However, the principal components can also be computed using the PCA-Seq method if segments of the original series are selected as objects. The matrix of Euclidean distances between the objects can be obtained using any method, which offers additional opportunities for time series analysis compared to the conventional SSA. In this study, the PCA-Seq method was used to analyze the dynamics of COVID-19 epidemic indicators.