TY - GEN
T1 - A sequential neighbor exploratory experimental design method for complex simulation metamodeling
AU - Lei, Yonglin
AU - Dong, Wei
AU - Zhu, Zhi
AU - Sarjoughian, Hessam
N1 - Publisher Copyright:
©2019 Society for Modeling & Simulation International (SCS).
PY - 2019
Y1 - 2019
N2 - Since complex simulation contains many factors and responses that interact in a nonlinear manner, it is important to use metamodeling for representing the causal relationships within the simulation models in a compact way. With the selected mathematical structure, the effectiveness of metamodeling depends on the comprehensiveness of the training data that is closely related to the size of the scenario space and available computing resources. Generally, sequential experimental design methods are more efficient than one-shot ones, but the later depend on the domain knowledge of the experimenters and are uneasy to be conducted automatically. This paper proposes a sequential neighbor exploratory experimental design (SNEED) method for metamodeling purpose. Through the peaks function example, we compare this new method to Latin hypercube with a support vector regression metamodel trained by their training data respectively. The result shows that under the same experiment sample count, the SNEED method produces better regression performance.
AB - Since complex simulation contains many factors and responses that interact in a nonlinear manner, it is important to use metamodeling for representing the causal relationships within the simulation models in a compact way. With the selected mathematical structure, the effectiveness of metamodeling depends on the comprehensiveness of the training data that is closely related to the size of the scenario space and available computing resources. Generally, sequential experimental design methods are more efficient than one-shot ones, but the later depend on the domain knowledge of the experimenters and are uneasy to be conducted automatically. This paper proposes a sequential neighbor exploratory experimental design (SNEED) method for metamodeling purpose. Through the peaks function example, we compare this new method to Latin hypercube with a support vector regression metamodel trained by their training data respectively. The result shows that under the same experiment sample count, the SNEED method produces better regression performance.
KW - K-d tree
KW - Latin hypercube
KW - Neighbor exploratory
KW - Sequential experimental design
KW - Simulation metamodel
UR - http://www.scopus.com/inward/record.url?scp=85073698474&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85073698474&partnerID=8YFLogxK
U2 - 10.23919/SpringSim.2019.8732886
DO - 10.23919/SpringSim.2019.8732886
M3 - Conference contribution
AN - SCOPUS:85073698474
T3 - Simulation Series
BT - Simulation Series
PB - The Society for Modeling and Simulation International
T2 - 2019 Theory of Modeling and Simulation, TMS 2019, Part of the 2019 Spring Simulation Multi-Conference, SpringSim 2019
Y2 - 29 April 2019 through 2 May 2019
ER -