анализ связанности

Purpose. The aim is to reveal the dependence of Granger causality results on chosen time scales of constructed empirical models in application to the task of investigation of evolution of coupling between brain areas during limbic seizures.

Transfer entropy is widely used to detect the directed coupling in oscillatory systems from their observed time series. The systematic error is detected, while estimating transfer entropy between nonlinear systems with K-nearest neighbours method.

Estimating coupling between systems of different nature is an urgent field of nonlinear dynamics method application. This work aims to compare classical linear Granger approach and its nonlinear analogues based on analysis of ethalon dynamical systems and neurophysiological data. The results achieved show nonlinear approach to be more sensitive, and so it is able to detect significant coupling, when linear one fails.

The task is considered of taking into account the multiple time scales of original time series, with these time series being used for Granger causality estimation. It is proposed to use the combination of prediction length and lag, different in value, that could be fruitful for comparatively short times series, e. g. of medical-biological nature. The automated methods are constructed to select lag and prediction length values. The proposed approach is tested on a set of examples – ethalon systems.