Step 4: A convergence check is performed on the updated response surface. If one or more pre-specified convergence criteria (see Section 3.1 for definition) are met, then an optimization search is performed on the final response surface. The simulation model is run with this optimal parameter set and the corresponding objective function value is recorded and checked against the best objective function value from the prior search. If it is the new objective function value is better than the previous one, then we accept this parameter set from the surrogate model as the optimal parameter set. Otherwise, new sets of adaptive samples are generated until the resulted response surface is acceptable.