Online Program Home
  My Program

All Times EDT

Abstract Details

Activity Number: 420 - Highlights from "Bayesian Analysis": Bayesian Methods for the Public Good
Type: Invited
Date/Time: Thursday, August 6, 2020 : 10:00 AM to 11:50 AM
Sponsor: International Society for Bayesian Analysis (ISBA)
Abstract #309238
Title: Efficient Bayesian Model Choice for Partially Observed Processes: With Application to an Experimental Transmission Study of an Infectious Disease
Author(s): S. E. F. Spencer* and Trevelyan J McKinley and Peter Neal and Andrew Conlan and Laurence Tiley
Companies: University of Warwick and University of Exeter and Lancaster University and University of Cambridge and University of Cambridge
Keywords: Bayesian model selection; epidemic models; partially observed processes; alive particle filter; particle MCMC; Approximate Bayesian Computation
Abstract:

Statistical models can provide key insights into the mechanisms that underlie the spread and persistence of infectious diseases, though their utility is linked to the ability to adequately calibrate these models to observed data. The fact that the underlying transmission models exhibit complex non-linear dynamics, coupled with practical constraints on observing key epidemiological events such as infections, makes robust inference challenging. Any solution to the inference problem requires the use of methodologies that can integrate over multiple hidden states and infer missing information. Simulation-based techniques such as Approximate Bayesian Computation (ABC) show great promise in this area, since they rely only on the development of suitable simulation models. In this talk we make some contributions towards improving the efficiency of ABC-based particle Markov chain Monte Carlo methods, and show the utility of these approaches for performing both inference and model comparison in a Bayesian framework. We illustrate these approaches on both simulated data and real data from an experimental transmission study of highly pathogenic avian influenza in genetically modified chickens.


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

Back to the full JSM 2020 program