|
Activity Number:
|
667
- Statistics, Science, and Society
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Thursday, August 2, 2018 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
IMS
|
|
Abstract #329208
|
Presentation
|
|
Title:
|
Branching Processes in Generalized Autoregressive Conditional Environments with Applications to Virus Outbreaks
|
|
Author(s):
|
Irene Hueter*
|
|
Companies:
|
Columbia University
|
|
Keywords:
|
GARCE processes;
Branching processes in random environment ;
Extinction ;
Phase transition;
Level shift ;
Ebola virus disease
|
|
Abstract:
|
Stochastic processes in autoregressive conditional environment offer a simple approach to analyze data modeled by a stochastic process in dynamic environment. Forecasts for future values of the random environment are easily generated, taking advantage of time series techniques. We illustrate key ideas in the GARCE branching processes - or branching processes in generalized autoregressive conditional environments - that allow for time-varying environments and instances of peak growth and near extinction-type rates as observed for viruses and infectious diseases. This novel approach to branching processes was successful to early detect the magnitude of the Ebola virus outbreak in West Africa when data became available in early 2014. With an added intervention effect, we examined whether the drastic control interventions initiated in late 2015 had a sufficient impact to stabilize and eventually end the outbreak and accurately anticipated the latter and its timing. We present stationarity conditions of the environment and results on the survival-extinction dichotomy, growth behavior, and phase transition. Intervention analyses of the Ebola outbreak data serve to illuminate the method.
|
Authors who are presenting talks have a * after their name.
Back to the full JSM 2018 program
|
