|
Activity Number:
|
510
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Wednesday, August 5, 2009 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Business and Economic Statistics Section
|
| Abstract - #304427 |
|
Title:
|
A Bayesian Approach to Seasonal Long Memory
|
|
Author(s):
|
Scott Holan*+ and Tucker S. McElroy
|
|
Companies:
|
University of Missouri-Columbia and U.S. Census Bureau
|
|
Address:
|
146 Middlebush Hall, Columbia , MO, 65211-6100,
|
|
Keywords:
|
Bayesian hierarchical model ; reversible jump MCMC ; seasonal long-memory
|
|
Abstract:
|
We develop a Bayesian procedure for analyzing time series that exhibit seasonal long memory. Specifically, we consider the Generalized Exponential Model (GEXP) to estimate the trend and/or seasonal long memory parameters of a stationary Gaussian time series. Estimation of the GEXP proceeds using a Bayesian hierarchical model. Further, we provide a brief description of a reversible jump Markov chain Monte Carlo algorithm for variable dimension estimation. Simulations and analysis of real time series data demonstrate the effectiveness of our approach.
|
