State space models are considered for observations which have non-Gaussian distributions. We obtain accurate approximations to the loglikelihood for such models by Monte Carlo simulation. Devices are introduced which improve the accuracy of the approximations and which increase computational efficiency. The loglikelihood function is maximised numerically to obtain estimates of the unknown hyperparameters. Standard errors of the estimates due to simulation are calculated. Details are given for the important special cases where the observations come from an exponential family distribution and where the observation equation is linear but the observation errors are non-Gaussian. The techniques are illustrated with a series for which the observations have a Poisson distribution and a series for which the observation have a t-distribution.
Biometrika is primarily a journal of statistics in which emphasis is placed on papers containing original theoretical contributions of direct or potential value in applications. From time to time, papers in bordering fields are published.
Oxford University Press is a department of the University of Oxford. It furthers the University's objective of excellence in research, scholarship, and education by publishing worldwide. OUP is the world's largest university press with the widest global presence. It currently publishes more than 6,000 new publications a year, has offices in around fifty countries, and employs more than 5,500 people worldwide. It has become familiar to millions through a diverse publishing program that includes scholarly works in all academic disciplines, bibles, music, school and college textbooks, business books, dictionaries and reference books, and academic journals.
This item is part of JSTOR collection
For terms and use, please refer to our Terms and Conditions
Biometrika
© 1997 Biometrika Trust
Request Permissions
