报告题目：Optimal Semi-supervised Subsampling via Predictive Inference
报告人：邹长亮 南开大学
报告时间：2022年12月20日(周二) 16:30-18:00
报告地点：中科院数学与系统科学研究院南楼N204；
腾讯会议 ID：375-8612-5504
内容摘要
In big data era, subsampling or sub-data selection techniques are often adopted to extract a fraction of informative individuals from the massive data. Existing subsampling algorithms focus mainly on obtaining a representative subset to achieve best estimation accuracy under a given class of models. In this talk, we consider a semi-supervised setting wherein a small or moderate sized “labeled” data is available in addition to a much larger sized “unlabeled” data. The goal is to sample from the unlabeled data with a give budget to obtain informative individuals that are characterized by their unobserved responses. I will introduce an optimal subsampling procedure that is able to maximize the diversity of the selected subsample and control false selection rate (FSR) simultaneously, allowing us to explore reliable information as much as possible.
主讲人简介
邹长亮，南开大学统计与数据科学学院教授。2008年于南开大学获博士学位，随后留校任教。主要从事统计学及其与数据科学领域的交叉研究和实际应用。研究兴趣包括：高维数据统计推断、大规模数据流分析、变点和异常点检测等，在Annals of Statistics、Biometrika、Journal of the American Statistical Association 、Mathematical Programming、Technometrics、IISE Transactions等统计学和工业工程领域期刊上发表论文几十篇，主持国家自然科学基金委中优青、杰青、重点项目、重大项目课题等。
