Fitting an Unobserved Component Time Series Model.

Abstract

The paper deals with a comparison of two methods for estimating a parametric model of a linear, Gaussian time series, based upon two different approximations to the likelihood function of the process. Discussed is the estimation of the parameters of Nerlove's unobserved-component model as presented in 'Analysis of Economic Time Series by Box-Jenkins and Related Techniques', through use of two approximations to the likelihood function of a Gaussian time series.

Document Details

Document Type
Technical Report
Publication Date
Jan 01, 1973
Accession Number
AD0760358

Entities

People

  • P. Bloomfield
  • T. Corwin

Organizations

  • Princeton University

Tags

Fields of Study

  • Mathematics

Readers

  • Statistical inference.