Variational Bayesian non-parametric inference for infectious disease models
James J. Hensman, Theodore Kypraios · Institution of Engineering and Technology eBooks · 2016
Chapter Contents: 9.1 Introduction 9.1.1 Infectious disease modelling 9.1.2 Why non-parametric inference? 9.1.3 Previous work 9.2 Background 9.2.1 Gaussian processes 9.2.2 Variational Bayes 9.3 Modelling framework 9.3.1 SIR model definition 9.3.2 Approximating the SIR model with a log Gaussian Cox process 9.3.3 Relaxing the parametric assumptions of the SIR model 9.3.4 Bayesian inference for an LGCP 9.3.5 Sparse variational approximations to GPs 9.4 Results 9.4.1 Dataset 1: Synthetic data from a homogeneously mixing mass-action SIR model 9.4.2 Dataset 2: Synthetic data from a seasonal SIR model 9.4.3 Application to the Abakaliki Smallpox data 9.5 Conclusions Acknowledgements References