Conditional autoregressive models for geographically sparse outcomes

Thursday, 25 September 2014
Alfred Centre, Seminar Rm 1, Level 5
Commercial Rd

Spatial epidemiology is the description and analysis of geographically indexed health data with respect to demographic, environmental, behavioral, socioeconomic, genetic, and infectious risk factors. Common familiar regression models are not sufficient to analyse such data, as they do not account for inherent spatial correlation in the outcomes and exposure variables. Sparse outcomes also add to the challenges in the analysis of such small-area datasets. Using a series of published papers, this talk will highlight one such Bayesian model, the Conditional Autoregressive (CAR) model, with practical applications to various health datasets from Australia and Singapore.

Dr Arul Earnest

Dr Arul Earnest

Swinburne University

Arul’s research interest lies in Bayesian hierarchical random effects modeling. He has developed and applied these models on a number of diseases in Singapore and Australia. For more than 15 years, Arul has provided consultative and collaborative methodological input to clinicians and hospital administrators, which has led to more than 120 publications in a variety of peer-reviewed international medical journals, numerous presentations, and several research awards. Arul has secured several recent grants, including a large Australian ARC grant, as well as an NMRC grant from Singapore.  Arul’s previous appointment was with Duke-NUS in Singapore as an Associate Professor and Director, Centre for Quantitative Medicine. He is currently with the Department of Statistics, Data Science and Epidemiology, School of Health Sciences. Swinburne University of Technology.