Imputation of missing observations for heavy tailed cyclostationary time series / Christiana Drake, Jacek Leśkow, Aldo M. Garay // W: Cyclostationarity : theory and methods. [T.] 2. Contributions to the ... więcej

Imputation of missing observations for heavy tailed cyclostationary time series / Christiana Drake, Jacek Leśkow, Aldo M. Garay // W: Cyclostationarity : theory and methods. [T.] 2. Contributions to the 7th workshop on cyclostationary systems and their applications, Grodek, Poland, 2014 / eds. Fakher Chaari, Jacek Leskow, Antonio Napolitano [et al.]. – Cham [etc.] : Springer, 2015. – (Applied Condition Monitoring, ISSN 2363-698X ; 3). – S. 179-187 : rys., tab., wykr. – Bibliogr. 9 poz. – doi: 10.1007/978-3-319-16330-7_9. – ISBN 978-3-319-16329-1

Periodically correlated (or cyclostationary) time series have many applications in engineering, environmental sciences, and economics. However, in many practical situations data that can be modelled with ... więcej

Periodically correlated (or cyclostationary) time series have many applications in engineering, environmental sciences, and economics. However, in many practical situations data that can be modelled with such time series is incomplete. An algorithm addressing imputation when some observations are missing at random has been developed in [2], where a version of the EM method had been employed. The mentioned algorithm was developed for AR(p)times series with Gaussian errors. In this paper we propose four algorithms for a K-dependent time series that come from the multivariate t-distribution.