时 间：2023年11月25日 09:00-11:30
地 点：普陀校区理科大楼A1514 腾讯会议ID：615-1716-9796
报告人：黄坚 香港理工大学教授
主持人：周勇 华东师范大学教授
摘 要：
Data representation learning is a crucial step in developing effective analysis models in statistics and machine learning. In particular, the goal of supervised representation learning is to construct effective data representations for prediction. Among all the characteristics of an ideal representation of high-dimensional complex data, sufficiency, low dimensionality, and disentanglement are some of the most essential ones. We propose a deep dimension reduction approach to learning representations with these characteristics. We formulate the representation learning task as that of finding a representation that minimizes an objective function characterizing conditional independence and promoting disentanglement. We estimate the target representation function nonparametrically using deep neural networks. We show that the estimated deep nonparametric representation is consistent in the sense that its excess risk converges to zero. Our extensive numerical experiments using simulated and real benchmark datasets demonstrate that the proposed methods have better performance than several existing dimension reduction methods and the standard deep learning models in the context of classification and regression. This is joint work with Yuling Jiao, Xu Liao, Jin Liu, and Zhou Yu.
报告人简介：
黄坚，香港理工大学应用数学系数据科学与分析讲座教授。在武汉大学获得数学学士和统计学硕士学位，在华盛顿大学（西雅图）获得统计学博士学位。曾任美国爱荷华大学统计学与精算学系和生物统计学系教授。其研究兴趣包括机器学习，高维统计，计算统计，生物统计学，和生物信息学。近年来主要致力于深度学习的研究,包括生成学习、表示学习，深度神经网络逼近理论、深度学习的理论分析、及其在数据科学中的应用。在包括《The Annals of Statistics》、《Journal of The American Statistical Association》，《Journal of Machine Learning Research》，《Biometrika》，《Conference on Neural Information Processing Systems》等国际统计学、机器学习杂志和会刊，和《Econometrika》，《Journal of Econometrics》等计量经济学杂志，以及《PNAS》，《American Journal of Human Genetics》，《Bioinformatics》,《Nucleic Acid Research》等杂志上发表学术论文150余篇。从2015年到2019年入选科睿唯安全球高被引学者榜，在数学领域里被引用最多的论文中排名前百分之一；并在2022,2023年被斯坦福大学列入全球前2% 被引用最多的科学家名单。是美国统计学会（ASA）和国际数理统计研究所（IMS）Fellow。