报告人：牛罡 高级研究科学家 日本理化学研究所
主持人：吕佳祺
报告时间：2025年10月30日（周四）上午10:00-11:00
报告地点：h片 九龙湖校区计算机楼513报告厅
报告摘要：Distribution shift (DS) may have two levels: the distribution itself changes, and the support (i.e., the set where the probability density is non-zero) also changes. When considering the support change between the training and test distributions, there can be four cases: (i) they exactly match; (ii) the training support is wider (and thus covers the test support); (iii) the test support is wider; (iv) they partially overlap. Existing methods are good at cases (i) and (ii), while cases (iii) and (iv) are more common nowadays but still under-explored. In this paper, we generalize importance weighting (IW), a golden solver for cases (i) and (ii), to a universal solver for all cases. Specifically, we first investigate why IW might fail in cases (iii) and (iv); based on the findings, we propose generalized IW (GIW) that could handle cases (iii) and (iv) and would reduce to IW in cases (i) and (ii). In GIW, the test support is split into an in-training (IT) part and an out-of-training (OOT) part, and the expected risk is decomposed into a weighted classification term over the IT part and a standard classification term over the OOT part, which guarantees the risk consistency of GIW. Then, the implementation of GIW consists of three components: (a) the split of validation data is carried out by the one-class support vector machine, (b) the first term of the empirical risk can be handled by any IW algorithm given training data and IT validation data, and (c) the second term just involves OOT validation data. Experiments demonstrate that GIW is a universal solver for DS problems, outperforming IW methods in cases (iii) and (iv).
报告人简介：Gang Niu is currently an indefinite-term senior research scientist at RIKEN Center for Advanced Intelligence Project. He received the PhD degree in computer science from Tokyo Institute of Technology in 2013. Before joining RIKEN, he was a senior software engineer at Baidu and then an assistant professor at the University of Tokyo. He joined RIKEN as a research scientist in 2018, and he was tenured in 2020 and promoted to senior research scientist in 2023. Gang published more than 130 journal articles and conference papers, including 42 ICML, 27 NeurIPS, and 17 ICLR (1 outstanding paper honorable mention) papers. He co-authored the book “Machine Learning from Weak Supervision: An Empirical Risk Minimization Approach” (the MIT Press). On the other hand, he served as a senior area chair/senior meta-reviewer 5 times and an area chair/meta-reviewer 27 times. He is also serving as an associate/action editor of IEEE TPAMI, MLJ, and TMLR. Moreover, he served as a publication chair for ICML 2022, and co-organized 18 workshops, 1 competition, and 3 tutorials. Last but not least, he is currently an IEEE senior member.
