A <scp>Bayesian</scp> approach to calibrate system dynamics models using <scp>Hamiltonian Monte Carlo</scp>

Model calibration is an essential test that dynamic hypotheses must pass in order to serve as tools for decision-making. In short, it the search a match between actual and simulated behaviours using parameter inference. Here, we approach such inference process from Bayesian perspective. Under this paradigm, provide statements about parameters (viewed random variables) data probabilistic terms. These stem posterior distribution whose solution often found via statistical simulation. However, uptake of these methods within system dynamics field has been somewhat limited, state-of-the-art algorithms have not explored. Therefore, introduce Hamiltonian Monte Carlo (HMC), efficient algorithm outperforms random-walk exploring complex spaces. We apply HMC calibrate SEIR model frame practical workflow. doing so, also recommend visualisation facilitate communication results. © 2021 The Authors. System Dynamics Review published by John Wiley & Sons Ltd on behalf Society.