Using Approximate Bayesian Computation (ABC) for parameter inference in SimInf [article]

Stefan Widgren
2022 Zenodo  
The objectives of OH-EJP FULL_FORCE WP5 (modelling) are to address: i) gaps in quantitative knowledge on the spread of AMR which will be essential to direct future focused research, ii) insight in the uncertainty around the effect of measures reducing AMR prevalence in the food production chains, and iii) identification of key elements in the production chains to mitigate the risk of human exposure. Estimating model parameters from observed data is a major challenge in stochastic modelling.
more » ... ver, it is a necessary and critical step to evaluate a model's explanatory power. Parameterization is preferably conducted within a Bayesian framework, and in this document, we will focus on Approximate Bayesian computation (ABC) (Toni et al., 2009), a simulation-based computational approach for parameterisation of complex problems, for example, in epidemiology (McKinley et el., 2018).
doi:10.5281/zenodo.7319680 fatcat:2flya62ccvholapusy3wfl2que