报告信息
报告题目：Optimal Composite Likelihood Estimation and Prediction for Large-scale Gaussian Process Models
报告时间：2023年9月8日 14:00-15:00
报告地点：必发9988集团 B205
邀请人： 徐照光 副教授 必发9988集团 大数据与智能决策研究中心
报告内容和摘要
Large-scale Gaussian process (GP) models are becoming increasingly important and popularly used in machine learning. However, the standard modeling method of GP models, which uses the maximum likelihood method and the best linear unbiased predictor, is designed to run on a single computer which often has limited computational power even when used in a supercomputing center. Therefore, a growing demand exists for approximate alternatives, such as the composite likelihood methods, that can use the power of multiple computers. However, these alternative methods in the literature offer limited options for practitioners because most methods focus more on computational efficiency than statistical efficiency. Limited accurate solutions to the parameter estimation and prediction of large-scale GP are provided in the literature for supercomputing practitioners. Therefore, this study develops an optimal composite likelihood (OCL) scheme that can minimize information loss in the parameter estimation and prediction of large-scale GP models through distributed computing. The proposed predictor, called the best linear unbiased block predictor (BLUBP), has the minimum prediction variance given the partitioned data. Numerical examples illustrate that both the proposed composite likelihood estimation and prediction methods exhibit more accurate performance than their traditional counterparts under various cases, and an extremely close approximation to the standard modeling method is observed.
报告人简介
李勇祥博士现为上海交通大学工业工程与管理系副教授，于2019年在香港城市大学数据学科学院取得哲学博士学位。李勇祥博士围绕复杂系统的质量与可靠性，研究机理与大数据联合驱动的复杂系统不确定性分析与量化，主要涉及复杂系统的试验设计、统计监测、智能诊断等。研究方向主要包括计算机实验设计与分析、统计与机器学习、统计质量控制、统计信号处理。代表性成果发表在相关领域国际高水平期刊《Technometrics》、《IEEE Transactions on Signal Processing》、《IISE Transactions》、《IEEE Transactions on Neural Networks and Learning Systems》、《IEEE Transactions on Industrial Electronics》等。李勇祥博士先后主持国家自然科学基金青年项目一项，上海市科技创新行动计划自然科学基金面上项目一项，并于2021年入选上海市浦江（A类）人才计划。
上一条：动态环境下竞争失效系统可靠性与维修性研究
下一条：Are Political Connections Really Conducive to Corporate Innovation? A Regression Discontinuity Design
【关闭】
