ON VECTOR AUTOCORRELATIONS AND GENERALIZED SECOND‐ORDER FUNCTIONS FOR TIME SERIES
Date
1994Source
Journal of Time Series AnalysisVolume
15Issue
3Pages
325-334Google Scholar check
Keyword(s):
Metadata
Show full item recordAbstract
Abstract. Some structural properties of certain vector generalizations of second‐order functions of a stationary stochastic process based on determinantial functions of autocovariances are discussed. In particular, a generalized autocovariance function which retains all properties of the ordinary autocovariance function is considered and the linear dependence structure of certain scalar stochastic processes associated with this function is investigated. Properties of the normalized function are discussed and a duality property is found, according to which this function also generalizes in a natural way the ordinary partial autocorrelation function of stochastic processes. Copyright © 1994, Wiley Blackwell. All rights reserved