Rainfall data analysis of Queensland using the Tweedie generalized linear model

Toshio Ohnishi; Peter K Dunn

Back

Rainfall data analysis of Queensland using the Tweedie generalized linear model

Journal article

Open access

Rainfall data analysis of Queensland using the Tweedie generalized linear model

Toshio Ohnishi and Peter K Dunn

RIMS Kokyuroku, Vol.1621, pp.135-152

2009

Files and links (2)

pdf

PDF - Published Version1.45 MBDownload View

Published Version PDF - Published Version Open Access

url

http://www.kurims.kyoto-u.ac.jp/~kyodo/kokyuroku/contents/pdf/1621-10.pdfView

Webpage

Abstract

Statistics

common slope

conjugate analysis

empirical Bayesian estimation

estimating function

generalized linear model

logarithmic link

Pythagorean relationship

posterior mode

Tweedie distribution

We investigate a logarithmic-link generalized linear model, whose underlying sam- pling density is in an exponential fainily distribution with power variance function. The multiple-strata case is studied with stratum-dependent intercepts and a common slope. We prove that there exists a conjugate prior density on the intercept parameter, and the conjugate analysis is discussed. An estimation procedure is given, which incIudes the optimal estimating function of the parameters other than the intercept, and an empirical Baysian estimation of the hyper-parameters of the prior density. As an example, rainfall data for Queensland, Australia, is analyzed.

Details

Title: Rainfall data analysis of Queensland using the Tweedie generalized linear model
Authors: Toshio Ohnishi (Author) - Institute of Statistical Mathematics, Japan
Peter K Dunn (Author) - University of the Sunshine Coast - Faculty of Science, Health and Education
Publication details: RIMS Kokyuroku, Vol.1621, pp.135-152
Publisher: Kyoto University, Research Institute for Mathematical Sciences
Date published: 2009
ISSN: 1880-2818
Organisation Unit: University of the Sunshine Coast, Queensland; School of Health and Sport Sciences - Legacy; School of Science, Technology and Engineering
Language: English
Record Identifier: 99450665002621
Output Type: Journal article

Metrics

30 File views/ downloads

375 Record Views