On Saturday October 29, Mr. Hola Kwame Adrakey, a PhD candidate in Statistics at Heriot-Watt University (UK) visited CROP and gave a talk on some of the research he is currently conducting. His research activities are focused on a range of areas of applied Statistics. One main topic of his interest is the development of Bayesian methods for model fitting in epidemiology, plant and animal diseases in particular – to better predict the risk of spread into disease-free areas. This work is part of an active collaboration with Chris-Gilligan and co-workers in plant Sciences department (Cambridge), Biomathematical and Statistics Scotland (BioSS), and Max Lau of Princeton University.
Mr. Adrakey began the presentation by introducing stochastic dynamical modelling in Bayesian framework. He explained why stochastic models are important in epidemiological modelling, before moving on to explain how data augmentation can facilitate the exploration of a posterior distribution.
Some illustrative examples on real-world data from some of his recent works on how MCMC and data augmentation could be valuable in targeting cost-effective control strategies to reducing the risk of propagation of a disease was discussed.