报告人：Liang Zhao (赵亮) 教授
报告题目：Machine Learning in Complex Networks
In the last years, an increasing interest in machine learning and data mining techniques using complex networks has been verified. This emergence is explained by the inherent advantages of network representation, which allows to capture spatial, topological, and functional relations of the data. In this talk, I firstly give an overview of complex network-based machine learning. Then, I introduce some specific techniques for supervised, unsupervised, and semi-supervised learning. Particularly, a particle competition model for both unsupervised and semi-supervised learning is described. Here, I also show how to deal with the problem of imperfect learning by adapting the particle competition system to withstand flawed training sets. Regarding supervised learning, a hybrid classification technique that combines both low and high orders of learning is introduced. The low-level term is implemented by traditional classification techniques, while the high-level term is realized by extracting features of the underlying network constructed from the input data. The general idea of the model is that the low-level term classifies test instances by their physical features, while the high-level term measures the compliance of test instances with the pattern formation of the data. We show that the high-level technique can realize classification according to the semantic meaning of the data.
Dr. Liang Zhao received the B.S. degree from Wuhan University, Wuhan, China, and the M.Sc. and Ph.D. degrees from the Aeronautic Institute of Technology, São Paulo, Brazil, in 1988, 1996, and 1998, respectively, all in computer science. He joined the University of São Paulo (USP) as a faculty member in 2000, where he is currently a Full Professor of the Department of Computing and Mathematics (DCM). From 2015 to 2017, he was the head of DCM. From 2003 to 2004, he was a Visiting Researcher with the Department of Mathematics, Arizona State University, Tempe, USA. His current research interests include artificial neural networks, machine learning, complex networks, pattern recognition, and bioinformatics. Dr. Zhao is a recipient of the Brazilian Research Productivity Fellowship. He was an Associate Editor of the IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS from 2009 to 2012 and he is currently an Associate Editor of the NEURAL NETWORKS. Dr. Zhao has published 1 monograph, 2 edited books, and about 200 journal and conference papers.