Ensemble Approach for Recovering Phase Synchronization From Time Series

Abstract

An ensemble approach is presented for the reconstruction of a phase synchronization diagram from time series. As an example system, we analyze a forced Colpitts oscillator to show that synchronization diagram reconstructed by a single nonlinear model depends sensitively upon the model parameters, which should be estimated with a considerable amount of care. This dependence can be crucial for a precise recovery of the synchronization phenomena. To overcome this weakness, an ensemble approach is introduced. Two types of techniques, namely, (I) ensemble regression and (II) ensemble classification, are developed to show that they provide much more robust and reliable reconstruction of the synchronization diagram compared to the conventional single modeling approach.

Detecting anomalous phase synchronization from time series

“…To overcome this problem, an ensemble technique is utilized. 31 In this technique, Q sets of different nonlinear models F 1,2 ͑i͒ ͑i =1, ... ,Q͒ are collected, each of which can be obtained by the same procedure as ͑B2͒ except that a different initial condition, which is generated randomly, is set for the parameter estimation. This results in Q sets of neural networks with the same architecture with different parameter values W. An ensemble average is then taken as F 1,2 = ͑1 / Q͚͒ i=1 Q F 1,2 ͑i͒ .…”

Predicting phase synchronization of non–phase-coherent chaos

“…It has been shown that the ensemble technique provides much more reliable modeling of nonlinear dynamics compared to the case of utilizing only a single model, whose results are rather sensitive to the optimized parameters [26]. (P4) CPR index is computed [18].…”