Abstract for: Optimization & Calibration in Vensim
In this hands-on workshop we will explore some of the new optimization features in Vensim. While the primary focus is on the mechanics, we will also touch on the philosophy, mathematics, and limitations of the methods. We will use a single boom-and-bust market model as the backdrop for our calibration and policy experiments. Topics introduced, time permitting, will include payoff weighting in calibration, including Bayesian prior terms, MCMC for confidence bounds and simulated annealing, and stochastic optimization.