Singular spectral analysis of correlation functions of nonequidistant time series

A. I. Stankevich

Samara State Aerospace University Named by Academician S. P. Korolyov, Samara

Abstract: Application of singular spectrum analysis to autocorrelations of unevelly spaced time series is considered. A significant improvement of autocorrelation estimation is shown by imitation modeling.

Keywords: unevenly spaced time series, stochastic processes, autocorrelation, singular spectrum analysis.

UDC: 681.518.3