Prediction of Disease Dynamics with Structure Knowledge of Human Contact Networks

Recent studies show that the degree distribution of realistic contact networks impacts the prediction accuracy for disease dynamics during an epidemic. Based on the surveillance data of the Ebola out break in 2014, not only the basic structural knowledge degree distribution but also another structural knowledge clustering, affect the prediction accuracy for disease dynamics, and their impacts are different. In this paper, combining degree distribution with clustering, we design an new algorithm to predict disease dynamics with the improved accuracy. Based on our extensive experiments, we find that the structural knowledge (degree distribution and clustering) of contact networks is helpful to improve the prediction accuracy for disease dynamics, as compared with the algorithm that just considers the degree distribution of contact networks.