Activity Number:
|
284
|
Type:
|
Contributed
|
Date/Time:
|
Wednesday, August 14, 2002 : 8:30 AM to 10:20 AM
|
Sponsor:
|
Section on Bayesian Stat. Sciences*
|
Abstract - #301524 |
Title:
|
Compositional Time Series Analysis Using Dynamic Linear Model
|
Author(s):
|
Amitabha Bhaumik*+ and Dipak Dey+ and Nalini Ravishanker
|
Affiliation(s):
|
University of Connecticut and University of Connecticut and University of Connecticut
|
Address:
|
U-4120, 215 Glenbrook Road, Storrs, Connecticut, 06269, USA U-4120, 215 Glenbrook Road, Storrs, Connecticut, 06269, USA
|
Keywords:
|
Kalman filtering ; Box-Cox transformation ; Multivariate-t distribution
|
Abstract:
|
Compositional data comprises of multivariate observations that are essentially proportions of a whole quantity and occurs frequently in disciplines such as economics, geology, and ecology. Multivariate statistical procedures available in the literature for analyzing such data often lead to incorrect conclusions since they ignore the inherent constrained nature of these observations as parts of a whole. In this article, new techniques to model compositional time series data are described in a Bayesian hierarchical framework employing modified dynamic linear models. The approach is illustrated on compositional time series on world motor vehicle production by Japan, the USA, and all other countries.
|
- The address information is for the authors that have a + after their name.
- Authors who are presenting talks have a * after their name.
Back to the full JSM 2002 program |