Online Program Home
  My Program

Abstract Details

Activity Number: 298 - Ecology and Environmental Policy
Type: Contributed
Date/Time: Tuesday, August 1, 2017 : 8:30 AM to 10:20 AM
Sponsor: Section on Statistics and the Environment
Abstract #323889 View Presentation
Title: A Bayesian Penalized Hidden Markov Model for Animal Interactions
Author(s): Meridith Bartley* and Ephraim M Hanks and David Hughes
Companies: Penn State University and The Pennsylvania State University and Penn State Univeristy
Keywords: Ecological Modeling ; HMM ; Bayesian Estimation ; Ant Trophallaxis ; Penalization ; Bayesian Ridge Regression
Abstract:

Interactions between social animals provide insights into the exchange and flow of nutrients, disease, and social contacts. We consider an analysis of trophallaxis interactions between carpenter ants (Camponotus pennsylvanicus) over 4 hours of second-by-second observations. The data show clear switches between fast and slow modes of trophallaxis; however, fitting a standard hidden Markov model (HMM) results in an estimated hidden state process that is overfit to this high resolution data, as the state process fluctuates an order of magnitude more quickly than is biologically reasonable. We propose a novel approach for penalized estimation of HMMs through a Bayesian ridge prior on the state transition rates This penalty induces smoothing, limiting the rate of state switching to ensure more biologically feasible results. We develop a Markov chain Monte Carlo algorithm to perform Bayesian inference based on discretized observations of the contact network.


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

Back to the full JSM 2017 program

 
 
Copyright © American Statistical Association