Data-driven mathematical modelling can enrich our understanding of infectious disease spread enormously. Deardon et al (2010) introduced a class of individual level disease transmission model allowing the incorporation of different individual level covariates such as spatial location, vaccination status, etc. However, with even a moderate number of covariates, fitting models can be problematic due to issues with MCMC convergence, for example. This can make model-choice/variable selection highly time-consuming and difficult. Here we consider methods for fitting spatial versions of individual-level disease transmission models when we have many potential covariates to include in the model. The aim is to enhance the predictive ability and interpretability of models, and ease computational burden when fitting the models to data. Here, we focus on the use of a lasso penalty in a two-stage variable screening method choose the best subsets of covariates to include in the final model.

R. Deardon, S. P. Brooks, B. T. Grenfell, M. J. Keeling, M. J. Tildesley, N. J. Savill, D. J. Shaw & M. E. J. Woolhouse (2010), “Inference for individual-level models of infectious diseases in large populations” in Statistica Sinica, 20(1), 239-261.