Modeling Dependence via Copula of Functionals of Fourier Coefficients

TL;DR

This paper introduces a copula-based method to characterize complex, non-linear dependence in stationary time series using Fourier coefficients, surpassing traditional correlation measures, and demonstrates its application on neural data.

Contribution

It develops a novel copula-based framework for analyzing complex dependence in time series, extending coherence with a flexible parametric approach and applying it to neural data.

Findings

Captures non-linear dependence missed by traditional measures

Demonstrates effectiveness on simulated non-linear data

Applied successfully to rat local field potential data

Abstract

The goal of this paper is to develop a measure for characterizing complex dependence between stationary time series that cannot be captured by traditional measures such as correlation and coherence. Our approach is to use copula models of functionals of the Fourier coefficients which is a generalization of coherence. Here, we use standard parametric copula models with a single parameter both from elliptical and Archimedean families. Our approach is to analyze changes in local field potentials in the rat cortex prior to and immediately following the onset of stroke. We present the necessary theoretical background, the multivariate models and an illustration of our methodology on these local field potential data. Simulations with non-linear dependent data show information that were missed by not taking into account dependence on specific frequencies. Moreover, these simulations demonstrate how our proposed method captures more complex non-linear dependence between time series. Finally, we illustrate our copula-based approach in the analysis of local field potentials of rats.