新闻中心
学术报告

学术报告:Machine-learning Based Multi-Fidelity Surrogate Modeling Methods

作者:   发布时间:2021-05-18


邀 请 人

刘力军

学科方向

数学

报告人

张超

地区(国内/国外、境内/境外)

国内

职称

副教授

报告题目

Machine-learning Based Multi-Fidelity Surrogate Modeling Methods

报告时间

2021/05/19   13:30-15:30

线上平台

腾讯会议

会议ID

221 704 340

会议密码

0519

会议网址

空白

工作单位

大连理工大学

研究领域

机器学习、随机矩阵、大型装备制造智能运维等

报告人简介

张超:副教授、博导。IEEE会员、IMLS会员,AAAI协会助理。主要研究方向为:人工智能、机器学习基础理论、大数据分析、数据驱动的智能装备运维管理等。发表SCI论文近40篇、EI论文10余篇,其中包括机器学习顶级期刊JMLRIEEE-TIT和顶级会议NIPSUAIAISTATS等。多次担任IJCAIAAAIICDMSDMCIKMICCV-SDLCV等人工智能与数据挖掘领域顶级学术会议委员。同年获批主持2项国家自然科学基金项目(青年、面上),主持国家重点研发计划课题1项。

报告内容简介

High-fidelity (HF) samples are accurate but are obtained at high cost. In contrast,low-fidelity (LF)samples are widely available but provide rough approximations. Multi-fidelity modeling aims to incorporate massive LF samples with a small amount of HF samples to develop a model for accurately approximating the HF responses to unseen inputs.In the literature,a main body of MFS models, under the assumption that there is a linear-trend relation between LF and HF data, are derived from the interpolation methods including the radial basis function method,the kriging method and the polynomial response surface (PRS) method. However,the linear-trend assumption will not always hold in practice, and thus the interpolation-based MFS models usually have a limited generality.Instead, some machine-learning methods (such as feed forward neural network and support vector regression) were used to handle MFS modeling tasks regardless of the relatedness between HF and LF data. Nevertheless,when the HF samples are rarely few, the training of machine-learning based MFS models is still challenged because the number of undetermined parameters is usually smaller than the HF-sample size.In this study, we propose two new machine-learning based MFS models: the hierarchical regression framework and the generative adversarial network. Compared with the classical MFS models,the two models have a better performance with a lower demand on HF-sample size and meanwhile without any assumption on the relation between HF and LF data.

 


下一条:中国科学院应用数学研究所黄飞敏研究员应邀访问我院