Inference for Computationally Intensive Space-Time Disease Models

• Activity Number: 461
• Type: Invited
• Date/Time: Wednesday, August 7, 2013 : 8:30 AM to 10:20 AM
• Sponsor: Section on Statistics and the Environment
• Author(s): Murali Haran*+ and Roman Jandarov
• Companies: The Pennsylvania State U. and University of Washington
• Keywords: infectious disease ; Markov chain Monte Carlo ; spatiotemporal models ; disease dynamics ; gravity models

Abstract

Contagious, acute, immunizing infections like measles and pertussis can exhibit spatiotemporal dynamics that depend on the nature of spatial contagion and spatiotemporal variations in population structure and demography. Metapopulation models for the dynamics of these diseases are often of interest as they provide a framework for characterizing their spread. Fitting these models to data, however, is often challenging. Computational issues may stem from having too many latent variables or from expensive likelihood function evaluations. Furthermore, standard likelihood-based inference often results in unreliable inference. I will describe a computationally efficient inferential approach that results in reliable parameter estimates. I will discuss the application of these methods to some real data examples.