calibration with AI and BayCANN

BayCANN is a project involving artificial intelligence (AI) and uses artificial neural networks to conduct Bayesian calibration on simulation models.

BayCANN uses AI to learn how the model produces output from a set of inputs. For that, it uses artificial neural networks as a “metamodel” of the model. Once the neural net “learns” the model structure, we use it to conduct Bayesian calibration.

BayCANN is extremely efficient, with a calibration for a complex model taking less than an hour on a standard computer.