Nonlinear parametric model for Granger causality of time series

TL;DR

This paper introduces a flexible nonlinear Granger causality model using radial basis functions, capable of capturing complex interactions in time series data, demonstrated through physiological and neural network examples.

Contribution

It extends previous models by allowing non-additive, fully expressive functions for causality analysis in time series.

Findings

Effective in physiological data analysis

Successfully models neural feedback loops

Demonstrates improved causality detection

Abstract

We generalize a previously proposed approach for nonlinear Granger causality of time series, based on radial basis function. The proposed model is not constrained to be additive in variables from the two time series and can approximate any function of these variables, still being suitable to evaluate causality. Usefulness of this measure of causality is shown in a physiological example and in the study of the feed-back loop in a model of excitatory and inhibitory neurons.