Related Experiment Videos
[Numerical procedure for the decomposition of biorhythmic processes into harmonic oscillating components]
Gegenbaurs Morphologisches Jahrbuch
|January 1, 1977
Summary
This study introduces a numerical method to decompose time series into periodic harmonic oscillations and stochastic components. The procedure enhances harmonic analysis and statistical significance testing for time series decomposition.
Area of Science:
- Time Series Analysis
- Signal Processing
- Mathematical Modeling
Context:
- Rhythmic processes are often modeled as a sum of harmonic oscillations and a stochastic component.
- Accurate decomposition is crucial for understanding underlying periodicities in data.
- Existing methods may lack robustness in identifying and quantifying these components.
Purpose:
- To present a novel numerical procedure for decomposing time series into periodic nonstochastic components (harmonic oscillations) and a stochastic component.
- To combine Harmonic Analysis with Fisher's significance test for robust component identification.
- To provide a method for parameter estimation and successive elimination of identified oscillations.
Summary:
- A numerical procedure is detailed for decomposing time series into inherent periodic nonstochastic components (harmonic oscillations) and a stochastic component.
- The method integrates Harmonic Analysis across varying data intervals with R.A. Fisher's significance test to identify the dominant spectral component.
- Parameter values are estimated, and identified oscillations are successively removed from the time series, allowing for robust decomposition.
Impact:
- Enables more accurate identification and quantification of periodic components in time series data.
- Provides a framework for combining different numerical procedures for complex time series decomposition, particularly for biorhythmic processes.
- Enhances the statistical significance assessment of decomposed harmonic oscillating components.