A comparative study of expected improvement-assisted global optimization with different surrogates

Wang, Hu; Ye, Fan; Li, Enying; Li, Guangyao

doi:10.1080/0305215x.2015.1115645

Cited by 24 publications

(15 citation statements)

References 40 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In addition, the ensembles created by Chaudhuri and Haftka (2014), Wang et al . (2016), Bhattacharjee et al . (2018), Lv et al .…”

Section: Resultsmentioning

confidence: 99%

“…Badhurshah and Samad (2015) find that using an ensemble of surrogate-assisted optimization methods and computational fluid dynamics analysis, the optimality, efficiency of the optimization process, and the robustness of the optimum solutions can be improved. Wang et al (2016) use an EoS for global optimization to enhance the convergence ratio of an uncertainty predictor. Samad et al (2006) evaluate the performances of ensembles of surrogates in a turbomachinery blade-shape optimization.…”

Section: Frame Of Referencementioning

confidence: 99%

See 1 more Smart Citation

Ensemble of surrogates and cross-validation for rapid and accurate predictions using small data sets

Alizadeh

Jia

Nellippallil

et al. 2019

AIEDAM

View full text Add to dashboard Cite

In engineering design, surrogate models are often used instead of costly computer simulations. Typically, a single surrogate model is selected based on the previous experience. We observe, based on an analysis of the published literature, that fitting an ensemble of surrogates (EoS) based on cross-validation errors is more accurate but requires more computational time. In this paper, we propose a method to build an EoS that is both accurate and less computationally expensive. In the proposed method, the EoS is a weighted average surrogate of response surface models, kriging, and radial basis functions based on overall cross-validation error. We demonstrate that created EoS is accurate than individual surrogates even when fewer data points are used, so computationally efficient with relatively insensitive predictions. We demonstrate the use of an EoS using hot rod rolling as an example. Finally, we include a rule-based template which can be used for other problems with similar requirements, for example, the computational time, required accuracy, and the size of the data.

show abstract

“…In addition, the ensembles created by Chaudhuri and Haftka (2014), Wang et al . (2016), Bhattacharjee et al . (2018), Lv et al .…”

Section: Resultsmentioning

confidence: 99%

Section: Frame Of Referencementioning

confidence: 99%

Ensemble of surrogates and cross-validation for rapid and accurate predictions using small data sets

Alizadeh

Jia

Nellippallil

et al. 2019

AIEDAM

View full text Add to dashboard Cite

show abstract

“…The terms l(.) and t(.) refer the linear and hyperbolic tangent activation functions (transfer function) of last and hidden neurons, respectively. The accurate prediction of the ANN is evaluated using the determination factor (R 2 ) 27 and the mean square error (MSE). 28…”

Section: Integrated Ann-imoco Methodsmentioning

confidence: 99%

Multidisciplinary design optimization of stiffened panels using collaborative optimization and artificial neural network

Chagraoui

Soula

2017

Proceedings of the Institution of Mechanical Engineers, Part C:

View full text Add to dashboard Cite

A new method for solving the multidisciplinary design optimization problems with a minimal computational effort is presented. The proposed methodology is based on the combination of artificial neural network model and Improved Multi-Objective Collaborative Optimization. In the artificial neural network–Improved Multi-Objective Collaborative Optimization scheme, the back-propagation algorithm is used for training the artificial neural network metamodel and the Non-dominated Sorting Genetic Algorithm-II is used to search a Pareto optimality set for the objective functions of stiffened panels. The artificial neural network–Improved Multi-Objective Collaborative Optimization algorithm aims firstly to decompose the global optimization problem hierarchically into optimization design problem at system level and several sub-problems at sub-system level and secondly to replace each optimization problem at the system and subsystem levels by artificial neural network model to limit the computational cost. To highlight the efficiency and effectiveness of the proposed artificial neural network–Improved Multi-Objective Collaborative Optimization method, mathematical and engineering examples are presented. Results obtained from the application of the artificial neural network–Improved Multi-Objective Collaborative Optimization approach to an optimization problem of a stiffened panel are compared with those obtained by traditional optimization without using prediction tools. The new method (artificial neural network–Improved Multi-Objective Collaborative Optimization) was proven to be superior to traditional optimization. These results have confirmed the efficiency and effectiveness of the artificial neural network–Improved Multi-Objective Collaborative Optimization method. In addition, it converges at faster rate than traditional optimization. The traditional optimization method converges within 7918 s, while artificial neural network–Improved Multi-Objective Collaborative Optimization requires only 42 s, clearly, the artificial neural network–Improved Multi-Objective Collaborative Optimization method is much more efficient.

show abstract

“…In order to determine which surrogate model is better fitted within an EI-based EGO approach, a comparative study is led in [28]. Support Vector Regression, Radial Basis Function, Kriging, linear Shepard and an ensemble surrogates made of these latter are integrated into EGO and compared on 2-D and 6-D test functions and on a 4-D engineering problem.…”

Section: Surrogate Building Using Kriging Modelsmentioning

confidence: 99%

“…These problems are manageable by surrogate-assisted methods since they generally involve computationally expensive simulators or learning models tedious to train. In [28], parametrization of a predictive bending model is tackled. The design characteristics of the object to test and the characteristics of the load are given as inputs to the model and a value representing the deformation is provided as output.…”

Section: Problem Descriptionmentioning

confidence: 99%

Parallel surrogate-assisted optimization: Batched Bayesian Neural Network-assisted GA versus q-EGO

Briffoteaux

Gobert

Ragonnet

et al. 2020

Swarm and Evolutionary Computation

View full text Add to dashboard Cite

Surrogate-based optimization is widely used to deal with long-running black-box simulation-based objective functions. Actually, the use of a surrogate model such as Kriging or Artificial Neural Network allows to reduce the number of calls to the CPU time-intensive simulator. Bayesian optimization uses the ability of surrogates to provide useful information to help guiding effectively the optimization process. In this paper, the Efficient Global Optimization (EGO) reference framework is challenged by a Bayesian Neural Network-assisted Genetic Algorithm, namely BNN-GA. The Bayesian Neural Network (BNN) surrogate is chosen for its ability to provide an uncertainty measure of the prediction that allows to compute the Expected Improvement of a candidate solution in order to improve the exploration of the objective space. BNN is also more reliable than Kriging models for high-dimensional problems and faster to set up thanks to its incremental training. In addition, we propose a batch-based approach for the parallelization of BNN-GA that is challenged by a parallel version of EGO, called q-EGO. Parallel computing is a highly important complementary way (to surrogates) to deal with the computational burden of simulation-based optimization. The comparison of the two parallel approaches is experimentally performed through several benchmark functions and two real-world problems within the scope of Tuberculosis Transmission Control (TBTC). The study presented in this paper proves that parallel batched BNN-GA is a viable alternative to q-EGO approaches being more suitable for high-dimensional problems, parallelization impact, bigger databases and moderate search budgets. Moreover, a significant improvement of the solutions is obtained for the two TBTC problems tackled.

show abstract

A comparative study of expected improvement-assisted global optimization with different surrogates

Cited by 24 publications

References 40 publications

Ensemble of surrogates and cross-validation for rapid and accurate predictions using small data sets

Ensemble of surrogates and cross-validation for rapid and accurate predictions using small data sets

Multidisciplinary design optimization of stiffened panels using collaborative optimization and artificial neural network

Parallel surrogate-assisted optimization: Batched Bayesian Neural Network-assisted GA versus q-EGO

Contact Info

Product

Resources

About