scientific article; zbMATH DE number 3898064

From MaRDI portal
Publication:3677005





[For the entire collection see Zbl 0547.00060.] The paper is devoted to some modern approaches to time series model identification and spectral estimation. The information divergence of spectral density functions and the maximum entropy characterization of AR spectra are reminded. As the model identification is concerned the author distinguishes the models with no memory (or white noise), short memory (or stationary models), long memory (or nonstationary models) transformable to short memory or representable as the sum of a long memory signal plus a short memory noise. Various information criteria of ARMA model identification for short memory time series are discussed. Finally, the broad array of various spectral estimators is briefly described.











This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3677005)