Surrogate-Assisted multi-objective optimization algorithms (Bayesian Optimization) for expensive or time-consuming problems. Expensive problems include optimization related to neural network architecture, computational fluid dynamics, network simulation, large scale (variables, objectives and constraints) optimization, real work phenomena or data-driven approach.
If you are using the code, please cite our paper:
"A Taxonomy for Metamodeling Frameworks for Evolutionary Multiobjective Optimization", IEEE Transactions on Evolutionary Computation, Kalyanmoy Deb ; Rayan Hussein ; Proteek Chandan Roy ; Gregorio Toscano-Pulido
@ARTICLE{8340780, author={K. {Deb} and R. {Hussein} and P. C. {Roy} and G. {Toscano-Pulido}}, journal={IEEE Transactions on Evolutionary Computation}, title={A Taxonomy for Metamodeling Frameworks for Evolutionary Multiobjective Optimization}, year={2019}, volume={23}, number={1}, pages={104-116}, keywords={evolutionary computation;optimisation;statistical analysis;many-objective optimization problems;constraint function;evolutionary multiobjective optimization;computationally expensive procedures;plausible metamodeling frameworks;Kriging metamodeling approach;Metamodeling;Optimization;Taxonomy;Linear programming;Computational modeling;Complexity theory;Clustering algorithms;Evolutionary multiobjective optimization (EMO);Kriging;metamodel;surrogate model;taxonomy}, doi={10.1109/TEVC.2018.2828091}, ISSN={1089-778X}, month={Feb},}