Testing for correlation structures in short-term variabilities with long-term trends of multivariate time series

Area of Science:

• Data analysis
• Time series analysis
• Statistical modeling

Background:

• Multivariate time series often contain complex correlation structures.
• Identifying short-term variabilities amidst long-term trends is challenging.
• Existing methods may not adequately address independence in irregular fluctuations.

Purpose of the Study:

• To present a new method for identifying correlation structures in irregular fluctuations of multivariate time series.
• To test the null hypothesis of independence among short-term variabilities.
• To provide a robust tool for analyzing time series data with trends.

Main Methods:

• The study adapts the small shuffle surrogate method.
• The approach focuses on detecting short-term correlation structures.
• The method is validated using both simulated and experimental time series data.

Main Results:

• The developed method successfully identifies correlation structures in irregular fluctuations.
• It effectively distinguishes between correlated and independent short-term variabilities.
• The technique performs well on diverse datasets, including those with long-term trends.

Conclusions:

• The proposed method offers a reliable way to analyze short-term correlations in multivariate time series.
• It provides a valuable tool for researchers dealing with complex, trended data.
• The findings support the utility of the small shuffle surrogate method for assessing independence in time series fluctuations.