Surrogate Model Selection for Design Space Approximation And Surrogatebased Optimization
Abstract
Surrogate models are used to map input data to output data when the actual relationship between the two is unknown or computationally expensive to evaluate for sensitivity analysis, uncertainty propagation and surrogate based optimization. This work evaluates the performance of eight surrogate modeling techniques for design space approximation and surrogate based optimization applications over a set of generated datasets with known characteristics. With this work, we aim to provide general rules for selecting an appropriate surrogate model form solely based on the characteristics of the data being modeled. Furthermore, the computational experiments revealed that, in general, multivariate adaptive regression spline models (MARS) and single hidden layer feed forward neural networks (ANN) yielded the most accurate predictions over the design space while Random Forest (RF) models most reliably identified the locations of the optimums when used for surrogate-based optimization.
- Authors:
-
- Auburn Univ., AL (United States)
- Publication Date:
- Research Org.:
- RAPID Manufacturing Institute (United States)
- Sponsoring Org.:
- USDOE Office of Energy Efficiency and Renewable Energy (EERE), Energy Efficiency Office. Advanced Manufacturing Office
- OSTI Identifier:
- 1647743
- Grant/Contract Number:
- EE0007888
- Resource Type:
- Accepted Manuscript
- Journal Name:
- Computer Aided Chemical Engineering
- Additional Journal Information:
- Journal Volume: 47; Journal ID: ISSN 1570-7946
- Publisher:
- Elsevier
- Country of Publication:
- United States
- Language:
- English
- Subject:
- 97 MATHEMATICS AND COMPUTING; Surrogate model; optimization; design space approximation; multivariate adaptive regression splines; random forests; artificial neural networks
Citation Formats
Williams, B. A., and Cremaschi, S. Surrogate Model Selection for Design Space Approximation And Surrogatebased Optimization. United States: N. p., 2019.
Web. doi:10.1016/b978-0-12-818597-1.50056-4.
Williams, B. A., & Cremaschi, S. Surrogate Model Selection for Design Space Approximation And Surrogatebased Optimization. United States. https://doi.org/10.1016/b978-0-12-818597-1.50056-4
Williams, B. A., and Cremaschi, S. Wed .
"Surrogate Model Selection for Design Space Approximation And Surrogatebased Optimization". United States. https://doi.org/10.1016/b978-0-12-818597-1.50056-4. https://www.osti.gov/servlets/purl/1647743.
@article{osti_1647743,
title = {Surrogate Model Selection for Design Space Approximation And Surrogatebased Optimization},
author = {Williams, B. A. and Cremaschi, S.},
abstractNote = {Surrogate models are used to map input data to output data when the actual relationship between the two is unknown or computationally expensive to evaluate for sensitivity analysis, uncertainty propagation and surrogate based optimization. This work evaluates the performance of eight surrogate modeling techniques for design space approximation and surrogate based optimization applications over a set of generated datasets with known characteristics. With this work, we aim to provide general rules for selecting an appropriate surrogate model form solely based on the characteristics of the data being modeled. Furthermore, the computational experiments revealed that, in general, multivariate adaptive regression spline models (MARS) and single hidden layer feed forward neural networks (ANN) yielded the most accurate predictions over the design space while Random Forest (RF) models most reliably identified the locations of the optimums when used for surrogate-based optimization.},
doi = {10.1016/b978-0-12-818597-1.50056-4},
journal = {Computer Aided Chemical Engineering},
number = ,
volume = 47,
place = {United States},
year = {Wed Jul 31 00:00:00 EDT 2019},
month = {Wed Jul 31 00:00:00 EDT 2019}
}
Web of Science