Fast algorithm of long-range cross-correlation analysis using Savitzky-Golay detrending filter and its application to biosignal analysis

To evaluate long-range cross-correlated behavior observed in bivariate time series, detrended cross-correlation analysis (DCCA) was proposed. In the DCCA, trends embedded in each time series are eliminated via piecewise least-squares polynomial fitting in the same way as the detrended fluctuation an...

Full description

Saved in:
Bibliographic Details
Published in2017 International Conference on Noise and Fluctuations (ICNF) pp. 1 - 4
Main Authors Tsujimoto, Yutaka, Miki, Yuki, Watanabe, Eiichi, Hayano, Junichiro, Yamamoto, Yoshiharu, Nomura, Taishin, Kiyono, Ken
Format Conference Proceeding
LanguageEnglish
Japanese
Published IEEE 01.06.2017
Subjects
Algorithm design and analysis
Filtering algorithms
Heart rate variability
Maximum likelihood detection
Nonlinear filters
Time series analysis
Online AccessGet full text

Cover

More Information
Summary:To evaluate long-range cross-correlated behavior observed in bivariate time series, detrended cross-correlation analysis (DCCA) was proposed. In the DCCA, trends embedded in each time series are eliminated via piecewise least-squares polynomial fitting in the same way as the detrended fluctuation analysis (DFA). In this paper, as an improved variant of DCCA, we propose a DCCA method using the Savitzky-Golay detrending filters and its fast implementation algorithm. In addition, as an application of our method, we analyze the cardiorespiratory interaction.
DOI:10.1109/ICNF.2017.7986015