学术报告第201942期

报告题目：迁移学习在序列标注中的应用

报 告 人：Joey Tianyi Zhou（周天异）（Scientist, Institute of High Performance Computing (IHPC) in Agency for Science, Technology and Research (A*STAR), Singapore.）

报告时间：2019年7月19日（星期五） 上午9:30

报告地点：suncitygroup太阳集团网址逸夫科技楼北楼903

Abstract: We propose a new architecture for addressing sequence labeling, termed Dual Adversarial Transfer Network (DATNet). Specifically, the proposed DATNet includes two variants, i.e., DATNet-F and DATNet-P, which are proposed to explore effective feature fusion between high and low resource. To address the noisy and imbalanced training data, we propose a novel Generalized Resource-Adversarial Discriminator (GRAD) and adopt adversarial training to boost model generalization. We investigate the effects of different components of DATNet across different domains and languages, and show that significant improvement can be obtained especially for low-resource data. Without augmenting any additional hand-crafted features, we achieve state-of-the-art performances on CoNLL, Twitter, PTB-WSJ, OntoNotes and Universal Dependencies with three popular sequence labeling tasks, i.e. Named entity recognition (NER), Part-of-Speech (POS) Tagging and Chunking.

报告人简历：Joey Tianyi Zhou（周天异）周天异博士目前在新加坡科技研究局,高性能研究所担任科学家职位,并且担任高性能计算团队负责人。他主持了新加坡多个与机器学习理论研究相关的大型研究课题，科研经费总计逾200万新元（1000万人民币），涉及领域包括城市安防,智能交通路线规划等。他曾经在美国硅谷的索尼研发中心担任高级研发工程师，并且负责公司无人车项目的视觉感知部分研发工作。他博士毕业于新加坡南洋理工大学, 并且发表30多篇国际顶级期刊和会议文章（CCF A类，中科院一区）其中包括JMLR, AIJ, TNNLS等等。他目前担任多个国际SCI期刊的副主编或者客座主编其中包括IEEE Access, IET image processing等。