MATLAB Toolbox for metamodeling and solving optimization problem. The function to surrogate and optimize can be evaluated from an external numerical simulator.
SBDOT Copyright (C) 2017 CEA - LETI, DOPT. Author: C. Durantin
Metamodeling features :
- Kriging (gaussian process based surrogate model)
- Radial Basis Function
- Co-kriging (multifidelity)
- CoRBF (multifidelity)
- BQQV Kriging (for qualitative variables)
Optimization features :
- CMAES (constrained single objective)
- NSGA-II (constrained multiobjective)
- MGDA (unconstrained multiobjective)
Adaptive Sampling strategy :
- Expected improvement (constrained global optimization)
- Gutmann criterion (constrained global optimization)
- Expected hypervolume improvement (multiobjective)
- Robust efficient global optimization