< Back to Article

Active Learning to Understand Infectious Disease Models and Improve Policy Making

Figure 1

Iterative active learning approach with a simulation model.

(1) A Latin hypercube design is used to make configuration files. (2) These configurations are used for the simulation model. (3) All input-response data are modeled with SR. (4) The surrogate models obtained with SR are used to achieve system understanding. The response prediction uncertainty can be used to adapt the experimental design (1) for the following modeling cycle.

Figure 1