Behavioral Study of Bayesian Neural Networks Under a Typical Surrogate Model-Assisted Evolutionary Search Framework
The machine learning method for surrogate modeling is a keystone in surrogate model-assisted evolutionary algorithms (SAEAs). The current arguably most widely used surrogate modeling methods in SAEAs are Gaussian process and radial basis function. This paper investigates the behavior of a machine le...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/11044325/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849425915680915456 |
|---|---|
| author | Yushi Liu Muhammad Ali Imran Masood Ur-Rehman Bo Liu |
| author_facet | Yushi Liu Muhammad Ali Imran Masood Ur-Rehman Bo Liu |
| author_sort | Yushi Liu |
| collection | DOAJ |
| description | The machine learning method for surrogate modeling is a keystone in surrogate model-assisted evolutionary algorithms (SAEAs). The current arguably most widely used surrogate modeling methods in SAEAs are Gaussian process and radial basis function. This paper investigates the behavior of a machine learning alternative, Bayesian neural network (BNN), which has seldom been paid attention to in SAEAs. Through empirical methods, the properties of the BNN model, its co-work with prescreening methods, and its comparison with other machine learning alternatives are investigated using a typical SAEA model management framework. Experimental results show the behavior of BNN models in terms of surrogate model prediction accuracy, the availability of prediction uncertainty estimation, and the training cost, demonstrating the potential of BNN being a competitive surrogate alternative in SAEAs. This is validated in various experiments, in which, under the same SAEA model management framework, SAEAs using BNN as the surrogate model save 83.1% and 42.5% modeling time compared to those using GP and the drop-out method, respectively. |
| format | Article |
| id | doaj-art-e79fdd98e1ab4eb6b944b5fc2fbc2639 |
| institution | Kabale University |
| issn | 2169-3536 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj-art-e79fdd98e1ab4eb6b944b5fc2fbc26392025-08-20T03:29:35ZengIEEEIEEE Access2169-35362025-01-011311181411183110.1109/ACCESS.2025.358115111044325Behavioral Study of Bayesian Neural Networks Under a Typical Surrogate Model-Assisted Evolutionary Search FrameworkYushi Liu0https://orcid.org/0000-0002-1546-6415Muhammad Ali Imran1https://orcid.org/0000-0003-4743-9136Masood Ur-Rehman2https://orcid.org/0000-0001-6926-7983Bo Liu3https://orcid.org/0000-0002-3093-4571School of Electronics and Communication Engineering, Sun Yat-sen University, Shenzhen, ChinaJames Watt School of Engineering, University of Glasgow, Glasgow, U.K.James Watt School of Engineering, University of Glasgow, Glasgow, U.K.James Watt School of Engineering, University of Glasgow, Glasgow, U.K.The machine learning method for surrogate modeling is a keystone in surrogate model-assisted evolutionary algorithms (SAEAs). The current arguably most widely used surrogate modeling methods in SAEAs are Gaussian process and radial basis function. This paper investigates the behavior of a machine learning alternative, Bayesian neural network (BNN), which has seldom been paid attention to in SAEAs. Through empirical methods, the properties of the BNN model, its co-work with prescreening methods, and its comparison with other machine learning alternatives are investigated using a typical SAEA model management framework. Experimental results show the behavior of BNN models in terms of surrogate model prediction accuracy, the availability of prediction uncertainty estimation, and the training cost, demonstrating the potential of BNN being a competitive surrogate alternative in SAEAs. This is validated in various experiments, in which, under the same SAEA model management framework, SAEAs using BNN as the surrogate model save 83.1% and 42.5% modeling time compared to those using GP and the drop-out method, respectively.https://ieeexplore.ieee.org/document/11044325/Bayesian neural networkcomputationally expensive optimizationsurrogate-model-assisted evolutionary algorithmsurrogate modeling |
| spellingShingle | Yushi Liu Muhammad Ali Imran Masood Ur-Rehman Bo Liu Behavioral Study of Bayesian Neural Networks Under a Typical Surrogate Model-Assisted Evolutionary Search Framework IEEE Access Bayesian neural network computationally expensive optimization surrogate-model-assisted evolutionary algorithm surrogate modeling |
| title | Behavioral Study of Bayesian Neural Networks Under a Typical Surrogate Model-Assisted Evolutionary Search Framework |
| title_full | Behavioral Study of Bayesian Neural Networks Under a Typical Surrogate Model-Assisted Evolutionary Search Framework |
| title_fullStr | Behavioral Study of Bayesian Neural Networks Under a Typical Surrogate Model-Assisted Evolutionary Search Framework |
| title_full_unstemmed | Behavioral Study of Bayesian Neural Networks Under a Typical Surrogate Model-Assisted Evolutionary Search Framework |
| title_short | Behavioral Study of Bayesian Neural Networks Under a Typical Surrogate Model-Assisted Evolutionary Search Framework |
| title_sort | behavioral study of bayesian neural networks under a typical surrogate model assisted evolutionary search framework |
| topic | Bayesian neural network computationally expensive optimization surrogate-model-assisted evolutionary algorithm surrogate modeling |
| url | https://ieeexplore.ieee.org/document/11044325/ |
| work_keys_str_mv | AT yushiliu behavioralstudyofbayesianneuralnetworksunderatypicalsurrogatemodelassistedevolutionarysearchframework AT muhammadaliimran behavioralstudyofbayesianneuralnetworksunderatypicalsurrogatemodelassistedevolutionarysearchframework AT masoodurrehman behavioralstudyofbayesianneuralnetworksunderatypicalsurrogatemodelassistedevolutionarysearchframework AT boliu behavioralstudyofbayesianneuralnetworksunderatypicalsurrogatemodelassistedevolutionarysearchframework |