Multi-stage design space reduction and metamodeling optimization method based on self-organizing maps and fuzzy clustering
作者：
Highlights：
• A multi-stage design space reduction methodology is proposed.
• Metamodel optimization is more cost effective to find the optimum design.
• Efficient metamodel can be built in the preliminary reduction space.
• The proper number of clusters is determined by utilizing cluster validity indices.
• Obtaining the accurate results of high nonlinear problems within a small region.
摘要
•A multi-stage design space reduction methodology is proposed.•Metamodel optimization is more cost effective to find the optimum design.•Efficient metamodel can be built in the preliminary reduction space.•The proper number of clusters is determined by utilizing cluster validity indices.•Obtaining the accurate results of high nonlinear problems within a small region.
论文关键词：Design space reduction,Self-organizing maps,Fuzzy clustering,GK cluster,Metamodeling
论文评审过程：Received 4 June 2014, Revised 25 October 2015, Accepted 26 October 2015, Available online 30 October 2015, Version of Record 18 November 2015.
论文官网地址：https://doi.org/10.1016/j.eswa.2015.10.033
