Guest lecture
Prof. Lei Xu
Department of Computer Science and Engineering,Chinese University of Hong Kong, PRC.
Bayesian Ying-Yang System and Theory
as A Unified Statistical
Learning Approach
Date Friday March 27th 1998 Place
Department of Computer Science
Teollisuuskatu 23 room B649 (6th floor)Time 13.15 - 15
Abstract
A unified statistical learning approach called Bayesian Ying-Yang (BYY) system and theory has been developed by the present author in recent years. It has been shown to function as a general theory for supervised learning, for unsupervised learning and its semi-unsupervised extension as well as for temporal modeling, for new theories and models on parameter learning, regularization, structural complexity selection, architecture design, and data smoothing.In this theory the major existing popular unsupervised learning approaches, supervised learning approaches and temporal modeling approaches are unified as special cases with new insights. Quite a number of new results are obtained, ranging from new techniques for regularization in parameter learning and new criteria for structural complexity selection (e.g., the number of clusters, the dimension of the subspace, the number of hidden units, the number of basis functions, the number of experts, and the number of states in state space models or HMM), to new algorithms for these existing models and new learning models. Experimental results on several applications (e.g., clustering, PCA, ICA, RBF nets, etc) will be demonstrated.
Biografia
Lei Xu joined the Department of Computer Science and Engineering at the CUHK in September 1993 as a senior lecturer and has been a professor since October 1996. He has also been a professor of the National Machine Perception Laboratory at Peking University since 1992, where he started as a postdoc at the Department of Mathematics in 1987 and then became one of ten exceptionally promoted young associate professors of the Peking University in 1988. During 1989-93, he worked as postdoc or senior research associate in several universities in Finland, Canada and USA, including Harvard and MIT.He is a past president of the Asian-Pacific Neural Networks Assembly, an IEEE Senior member, and an associate editor for six renown international journals on Neural Networks. He has published over 180 papers on neural networks, computer vision & pattern recognition, signal processing, and AI, given over ten keynote/invited/tutorial talks as well as served as program committee member and session co-chairs at international major NN conferences in recent years, including WCNN, IEEE-ICNN, ENNS-ICANN, ICONIP, IJCNN, NNCM, NIPS. Also, he was co-chair of the ICONIP'96 program committee and NIPS Workshops. He has received several Chinese national prestigious academic awards and also some international awards, and he is listed in several major Who's Who and the First Five Hundreds publications by CIBC, ABI and Marquis Who's Who.