报告题目：Gaussian Processes for Spatial Statistical and Machine Learning

报告时间：2024.06.30   10:00-11:00

报  告 人 ：Prof. Hao Zhang   Michigan State University 

报告地点：理学院东北楼二楼报告厅（209）

Abstract：Gaussian processes serve as a robust modeling framework, finding extensive use across various disciplines within statistical and machine learning applications. The kernel, or covariance matrix, is a pivotal component in modeling Gaussian processes, aiding in estimation and prediction. However, as the sample size expands, the kernel matrix tends to become ill-conditioned, necessitating approximation strategies for Gaussian likelihood or spatial prediction. In this talk, I will survey an array of such approximation methods and share recent findings in the field.