Hierarchical regression framework for multi-fidelity modeling
作者：
Highlights：
• The models developed under the framework outperform the existing ones in most cases.
• There is no specific assumption on the sample distribution or the sample structure.
• The computational cost is comparable to that of the state-of-art models.
• It has a high robustness for varying sample size, especially for very few HF samples.
• The framework has a high applicability in practice.
摘要
•The models developed under the framework outperform the existing ones in most cases.•There is no specific assumption on the sample distribution or the sample structure.•The computational cost is comparable to that of the state-of-art models.•It has a high robustness for varying sample size, especially for very few HF samples.•The framework has a high applicability in practice.
论文关键词：Bi-fidelity modeling,Multi-fidelity modeling,Hierarchical regressor,Machine learning,Recursive method
论文评审过程：Received 26 February 2020, Revised 5 October 2020, Accepted 30 October 2020, Available online 10 November 2020, Version of Record 24 December 2020.
论文官网地址：https://doi.org/10.1016/j.knosys.2020.106587
