Synchronization approach to analysis of multivariate data

点击翻译按钮获取中文摘要

The aim of this project is to develop and elaborate modern techniques of nonlinear data analysis capable of detecting time-varying weak interaction between irregular (noisy, chaotic, or both) systems from the signals registered at their outputs. For this purpose we use our recent achievements in synchronization theory that resulted in the development of a synchronization approach to data analysis. We mainly intend to study criteria of statistical significance of the synchronization analysis, extension of the technique to the case of multivariate data and spatially-extended systems, as well as evaluation of efficiency of this method with respect to traditional techniques. Special attention will be paid to the quantification of intensity and directionality of interaction. We will also build up a software tool box of data analysis techniques including implementation of newly developed as well as of standard algorithms; it will be available for other projects in this special research programme.