|
Activity Number:
|
222
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Tuesday, August 10, 2004 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Biometrics Section
|
| Abstract - #301942 |
|
Title:
|
A Generalized Open-loop Threshold Model for Non-normal Nonlinear Time Series: Plague in Kazakhstan as in Illustration
|
|
Author(s):
|
Noelle I. Samia*+ and Kung-sik Chan and Nils C. Stenseth
|
|
Companies:
|
University of Iowa and University of Iowa and University of Oslo
|
|
Address:
|
Dept. of Statistics and Actuarial Science, Iowa City, IA, 52242,
|
|
Keywords:
|
binomial distribution ; delay ; epizootic events ; exponential family ; stochastic regression
|
|
Abstract:
|
The Open-loop Threshold model is essentially a piecewise linear stochastic regression model, with the errors often assumed to be normal. We generalize the Open-loop Threshold model for time series regressions with non-normal errors. Specifically, it is assumed that the response variables conditionally belong to an exponential family, whose canonical parameters are linked to some piecewise linear stochastic regression function. We study the specific case where the response variables are conditionally Binomially distributed. Some probabilistic properties of the model and the large-sample properties of an estimation scheme will be discussed. Our approach is motivated by the need for modeling nonlinearity in serially correlated epizootic events. Data coming from monitoring conducted in a natural plague focus in Kazakhstan are used to illustrate this model by obtaining biologically meaningful conclusions regarding the threshold relationship between prevalence of plague and some covariates including past abundance of gerbils and other meteorological variables.
|
