|
Activity Number:
|
192
- Contributed Poster Presentations: Section on Statistics and the Environment
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, July 30, 2018 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Statistics and the Environment
|
|
Abstract #330221
|
|
|
Title:
|
Bayesian Bivariate Extreme Value Analysis with Application in Environmental Statistics
|
|
Author(s):
|
Yuan Tian* and Brian Reich
|
|
Companies:
|
North Carolina State University and North Carolina State University
|
|
Keywords:
|
Bivariate extreme value;
Generalized Pareto distribution;
Tail dependence
|
|
Abstract:
|
Extreme value analysis is widely used in environmental statistics. For example, to estimate the probability of 100-year flood. Generalized Pareto distribution is conventionally used to model the extreme points over a predetermined threshold. However, threshold selection can be a challenge in the application. In this project, we proposed a novel bivariate extreme value model without selecting threshold. The new model holds nice tail dependence properties. Tail behavior of marginal distribution is studied as well. The model fitting procedure and the flexibility are illustrated with a simulation study. An application to environmental data is presented at the end.
|
Authors who are presenting talks have a * after their name.
