Kriging with composite kernel learning for surrogate modeling in computer experiments

Pramudita Satria Palar, Koji Shimoyama

Research output: Chapter in Book/Report/Conference proceedingConference contribution

3 Citations (Scopus)


We introduce a composite kernel learning (CKL) technique to construct new kernel (covariance) functions for efficient surrogate modeling in engineering design. The CKL technique works by constructing a new kernel from the weighted sum combination of existing kernels where the weights and hyperparameters of Kriging are simultaneously trained. By deploying Kriging with CKL, we aim to better capture the input-output relationship of black-box functions in engineering design problems through the construction of more suitable kernels. The effectiveness of CKL-Kriging is demonstrated on two computational fluid dynamics-based problems, that is, an axial transonic rotor problem and a centrifugal diffuser design problem. The results show that the Kriging with CKL generally outperforms or performs similarly to the best performing kernel on the two test problems. In light of the results, we infer that Kriging with CKL is a promising surrogate modeling technique to be deployed for general engineering design problems.

Original languageEnglish
Title of host publicationAIAA Scitech 2019 Forum
PublisherAmerican Institute of Aeronautics and Astronautics Inc, AIAA
ISBN (Print)9781624105784
Publication statusPublished - 2019 Jan 1
EventAIAA Scitech Forum, 2019 - San Diego, United States
Duration: 2019 Jan 72019 Jan 11

Publication series

NameAIAA Scitech 2019 Forum


ConferenceAIAA Scitech Forum, 2019
Country/TerritoryUnited States
CitySan Diego

ASJC Scopus subject areas

  • Aerospace Engineering


Dive into the research topics of 'Kriging with composite kernel learning for surrogate modeling in computer experiments'. Together they form a unique fingerprint.

Cite this