Heterogeneous Transfer Learning with Real-world Applications
Qiang Yang, Hong Kong University of Science and Technology (qyang[at]cse.ust.hk)
In many real-world machine learning and data mining applications, we often face the problem where the training data are scarce in the feature space of interest, but much data are available in other feature spaces. Many existing learning techniques cannot make use of these auxiliary data, because these algorithms are based on the assumption that the training and test data must come from the same distribution and feature spaces. When this assumption does not hold, we have to seek novel techniques for ‘transferring’ the knowledge from one feature space to another. In this talk, I will present our recent works on heterogeneous transfer learning. I will describe how to identify the common parts of different feature spaces and learn a bridge between them to improve the learning performance in target task domains. I will also present several interesting applications of heterogeneous transfer learning, such as image clustering and classification, cross-domain classification and collaborative filtering.
Qiang Yang is a professor in the Department of Computer Science and Engineering, Hong Kong University of Science and Technology. His research interests are artificial intelligence, including automated planning, machine learning and data mining. He graduated from Peking University in 1982 with BSc. in Astrophysics, and obtained his MSc. degrees in Astrophysics and Computer Science from the University of Maryland, College Park in 1985 and 1987, respectively. He obtained his PhD in Computer Science from the University of Maryland, College Park in 1989. He was an assistant/associate professor at the University of Waterloo between 1989 and 1995, and a professor and NSERC Industrial Research Chair at Simon Fraser University in Canada from 1995 to 2001.
Qiang Yang has been active in research on artificial intelligence planning, machine learning and data mining. His research teams won the 2004 and 2005 ACM KDDCUP international competitions on data mining. He has been on several editorial boards of international journals, including IEEE Intelligent Systems, IEEE Transactions on Knowledge and Data Engineering and Web Intelligence. He has been an organizer for several international conferences in AI and data mining, including being the conference co-chair for ACM IUI 2010 and ICCBR 2001, program co-chair for PRICAI 2006 and PAKDD 2007, workshop chair for ACM KDD 2007, AAAI tutorial chair for AAAI 2005 and 2006, data mining contest chair for IEEE ICDM 2007 and 2009, and vice chair for ICDM 2006 and CIKM 2009. He is a fellow of IEEE and a member of AAAI and ACM. His home page is at http://www.cse.ust.hk/~qyang.
Discourse - Early problems, current successes, future challenges
Bonnie Webber, University of Edinburgh, UK (bonnie.webber[at]ed.ac.uk)
will look back through nearly forty years of computational research on discourse, noting some problems (such as context-dependence and inference) that were identified early on as a hindrance to further progress, some admirable successes that we have achieved so far in the development of algorithms and resources, and some challenges that we may want to (or that we may have to!) take up in the future, with particular attention to problems of data annotation and genre dependence.
Bonnie Webber was a researcher at Bolt Beranek and Newman while working on the PhD she received from Harvard University in 1978. She then taught in the Department of Computer and Information Science at the University of Pennsylvania for 20 years before joining the School of Informatics at the University of Edinburgh. Known for research on discourse and on question answering, she is a Past President of the Association for Computational Linguistics, co-developer (with Aravind Joshi, Rashmi Prasad, Alan Lee and Eleni Miltsakaki) of the Penn Discourse TreeBank, and co-editor (with Annie Zaenen and Martha Palmer) of the journal, Linguistic Issues in Language Technology.