Combining Global and Local Surrogate Models to Accelerate Evolutionary Optimization Export

@article{citeulike:3646470, abstract = {In this paper, we present a novel surrogate-assisted evolutionary optimization framework for solving computationally expensive problems. The proposed framework uses computationally cheap hierarchical surrogate models constructed through online learning to replace the exact computationally expensive objective functions during evolutionary search. At the first level, the framework employs a data-parallel Gaussian process based global surrogate model to filter the evolutionary algorithm (EA) population of promising individuals. Subsequently, these potential individuals undergo a memetic search in the form of Lamarckian learning at the second level. The Lamarckian evolution involves a trust-region enabled gradient-based search strategy that employs radial basis function local surrogate models to accelerate convergence. Numerical results are presented on a series of benchmark test functions and on an aerodynamic shape design problem. The results obtained suggest that the proposed optimization framework converges to good designs on a limited computational budget. Furthermore, it is shown that the new algorithm gives significant savings in computational cost when compared to the traditional evolutionary algorithm and other surrogate assisted optimization frameworks}, author = {Zhou, Zongzhao and Ong, Yew S. and Nair, Prasanth B. and Keane, Andy J. and Lum, Kai Y.}, booktitle = {Systems, Man, and Cybernetics, Part C: Applications and Reviews, IEEE Transactions on}, citeulike-article-id = {3646470}, citeulike-linkout-0 = {http://dx.doi.org/10.1109/TSMCC.2005.855506}, citeulike-linkout-1 = {http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=4033013}, doi = {10.1109/TSMCC.2005.855506}, journal = {Systems, Man, and Cybernetics, Part C: Applications and Reviews, IEEE Transactions on}, keywords = {evolutionary-computation, fitness, gaussian, optimization}, number = {1}, pages = {66--76}, posted-at = {2008-12-03 11:19:57}, priority = {2}, title = {Combining Global and Local Surrogate Models to Accelerate Evolutionary Optimization}, url = {http://dx.doi.org/10.1109/TSMCC.2005.855506}, volume = {37}, year = {2007} }

TY - JOUR ID - citeulike:3646470 L3 - citeulike-article-id:3646470 N2 - In this paper, we present a novel surrogate-assisted evolutionary optimization framework for solving computationally expensive problems. The proposed framework uses computationally cheap hierarchical surrogate models constructed through online learning to replace the exact computationally expensive objective functions during evolutionary search. At the first level, the framework employs a data-parallel Gaussian process based global surrogate model to filter the evolutionary algorithm (EA) population of promising individuals. Subsequently, these potential individuals undergo a memetic search in the form of Lamarckian learning at the second level. The Lamarckian evolution involves a trust-region enabled gradient-based search strategy that employs radial basis function local surrogate models to accelerate convergence. Numerical results are presented on a series of benchmark test functions and on an aerodynamic shape design problem. The results obtained suggest that the proposed optimization framework converges to good designs on a limited computational budget. Furthermore, it is shown that the new algorithm gives significant savings in computational cost when compared to the traditional evolutionary algorithm and other surrogate assisted optimization frameworks IS - 1 JF - Systems, Man, and Cybernetics, Part C: Applications and Reviews, IEEE Transactions on EP - 76 TI - Combining Global and Local Surrogate Models to Accelerate Evolutionary Optimization VL - 37 T2 - Systems, Man, and Cybernetics, Part C: Applications and Reviews, IEEE Transactions on SP - 66 KW - evolutionary-computation KW - fitness KW - gaussian KW - optimization AU - Zhou, Zongzhao AU - Ong, Yew AU - Nair, Prasanth AU - Keane, Andy AU - Lum, Kai PY - 2007/// UR - http://dx.doi.org/10.1109/TSMCC.2005.855506 ER -