Using Dynamic Bayesian Network (DBN) for Evaluation

dynamic Bayesian network to evaluate bovine tuberculosis eradication policy and risk factors in England's cattle farms 2008 to 2015

Data are available publicly as secondary data in Quarterly TB in cattle in Great Britain statistical notice (data to March 2018). Data are available for high-risk, edge-risk and low-risk areas of England. In our model nodes are elicited from literature and available documents regarding the bTB policy published by Government. Edges which show the level of effectiveness between the different nodes and show causal relationships between them are elicited from the domain experts and the available literature and the prior probabilities were elicited from the available literature and data. Then, DBN updated the prior probabilities every time that new information (evidence) was added to the network.