Online Program Home
My Program

Abstract Details

Activity Number: 123
Type: Contributed
Date/Time: Monday, August 1, 2016 : 8:30 AM to 10:20 AM
Sponsor: Section on Statistics and the Environment
Abstract #321252 View Presentation
Title: Modeling Short-Term Population Dynamics with Unobserved Latent Stages
Author(s): Gabriel Demuth* and Philip Dixon
Companies: Iowa State University and Iowa State University
Keywords: Population Dynamics ; Matrix Models ; Markov Process ; Latent Variables
Abstract:

Stage structured matrix models provide an effective tool for modeling long-term population dynamics. However it can be difficult to capture short term transient behavior, particularly if there is heterogeneity within the stages observed in data. We present a method for introducing unobserved latent population stages as a method for increasing model richness, and accounting for unobserved heterogeneity. Latent stages may have their own dynamics, and are related to data by requiring them to sum to the mean of observed stages. This allows for the creation of flexible models for complex population dynamics that fit within the stage structured matrix model structure, and are therefore simple to implement and interpret. The method is applied to data on spore development in the common fungus Colletotricum acutatum under laboratory settings. While simple models based on observed stages fail to reflect short term trends, the latent stage model successfully captures the short and long term dynamics of spore development.


Authors who are presenting talks have a * after their name.

Back to the full JSM 2016 program

 
 
Copyright © American Statistical Association