Model-free measure of coupling from embedding principle

TL;DR

This paper introduces a model-free, embedding-based measure of coupling between dynamical variables that is robust to noise and does not require density estimation, demonstrated on financial time series.

Contribution

It presents a novel coupling measure that is mathematically simple, does not assume dynamics, and is applicable to high-dimensional, noisy data.

Findings

Effective in detecting coupling in complex financial data

Robust to noise and high-dimensionality

Provides strict asymptotic bounds

Abstract

A model-free measure of coupling between dynamical variables is built from time series embedding principle. The approach described does not require a mathematical form for the dynamics to be assumed. The approach also does not require density estimation which is an intractable problem in high dimensions. The measure has strict asymptotic bounds and is robust to noise. The proposed approach is used to demonstrate coupling between complex time series from the finance world.