Dynamic Bayesian models for vector time series analysis & forecasting

Doctoral thesis English OPEN
Barbosa, Emanuel Pimentel
  • Subject: QA

This thesis considers the Bayesian analysis of general multivariate DLM's (Dynamic Linear\ud Models) for vector time series forecasting where the observational variance matrices are\ud unknown. This extends considerably some previous work based on conjugate analysis for\ud a special sub—class of vector DLM's where all marginal univariate models follow the same\ud structure.\ud The new methods developed in this thesis, are shown to have a better performance than\ud other competing approaches to vector DLM analysis, as for instance, the one based on the\ud Student t filter.\ud Practical aspects of implementation of the new methods, as well as some theoretical properties\ud are discussed, further model extensions are considered, including non—linear models\ud and some applications with real and simulated data are provided.
  • References (2)

    i) Observation Equation : yt = FtT .tlt + vt, vt "-' N (0;Vt) ii) System Equation : et = G. 2t _ i +tut , tut---• N (0; W) Hartigan,J.A.(1969) . Linear Bayes methods . J. Roy. Statist. Soc. , B , 31 , 446-454.

    West,M.(1981) . Robust sequential approximate Bayesian estimation. J. Roy. Statist. Soc. - B , 43 , 157-166.

  • Metrics
    views in OpenAIRE
    views in local repository
    downloads in local repository

    The information is available from the following content providers:

    From Number Of Views Number Of Downloads
    Warwick Research Archives Portal Repository - IRUS-UK 0 34
Share - Bookmark