Methods
Knowledge-guided generative surrogate modeling for high-dimensional design optimization under scarce data
- Context
- Surrogate models replace expensive simulations during design optimization.
- Gap
- High-dimensional design spaces combined with scarce training data make purely data-driven surrogates unreliable.
- Research goal
- Guide generative surrogate models with domain knowledge so that design optimization remains effective under scarce data.
This research theme investigates how domain knowledge can guide generative surrogate models in high-dimensional design optimization when data are scarce.
Selected studies
ReferencesGenerative surrogate modeling guided by domain knowledge
A generative surrogate modeling approach incorporates domain knowledge to support high-dimensional design optimization when only scarce data are available.
Primary paper
Bingran Wang, Seongha Jeong, Sebastiaan P. C. van Schie, Dongyeon Han, Jaeho Min, John T. Hwang. Knowledge-Guided Generative Surrogate Modeling for High-Dimensional Design Optimization Under Scarce Data. Journal of Computing and Information Science in Engineering, 2026.
DOIPDF
@article{wang2026knowledge,
author = {Bingran Wang and Seongha Jeong and Sebastiaan P. C. van Schie and Dongyeon Han and Jaeho Min and John T. Hwang},
title = {Knowledge-Guided Generative Surrogate Modeling for High-Dimensional Design Optimization Under Scarce Data},
journal = {Journal of Computing and Information Science in Engineering},
year = {2026},
volume = {26},
number = {7},
doi = {10.1115/1.4070934},
url = {https://doi.org/10.1115/1.4070934},
note = {Metadata verification: Crossref title}
}References
- Bingran Wang, Seongha Jeong, Sebastiaan P. C. van Schie, Dongyeon Han, Jaeho Min, John T. Hwang. Knowledge-Guided Generative Surrogate Modeling for High-Dimensional Design Optimization Under Scarce Data. Journal of Computing and Information Science in Engineering, 2026.DOIPDF
@article{wang2026knowledge, author = {Bingran Wang and Seongha Jeong and Sebastiaan P. C. van Schie and Dongyeon Han and Jaeho Min and John T. Hwang}, title = {Knowledge-Guided Generative Surrogate Modeling for High-Dimensional Design Optimization Under Scarce Data}, journal = {Journal of Computing and Information Science in Engineering}, year = {2026}, volume = {26}, number = {7}, doi = {10.1115/1.4070934}, url = {https://doi.org/10.1115/1.4070934}, note = {Metadata verification: Crossref title} }
Research connections