Computing That Serves

Transfer Learning Via Advice Taking


Thursday, March 5, 2009 - 10:00am


Jude Shavlik
Professor, Departments of Computer Sciences and Biostatistics & Medical Informatics, University of Wisconsin-Madison

Transfer learning is the use of knowledge learned in a source task to improve learning in a related target task. An important question is the mechanism by which learned knowledge in one task is converted to a form usable when learning the second task.  Our approach is to follow the "theory refinement" approach to machine learning.  In theory refinement, which is also called "advice taking," approximately correct domain knowledge ("advice") is given to a learning algorithm that then uses training examples to correct and extend the provided domain knowledge.  We have developed several algorithms that convert what was learned in the source task into advice about the target task.  We have also developed an approach called Knowledge-Based Support Vector Machines that accepts and refines advice.  In addition to explaining these algorithms, we present empirical results using the RoboCup soccer simulator.


Jude Shavlik is a Professor of Computer Sciences and of Biostatistics and Medical Informatics at the University of Wisconsin - Madison, and is a Fellow of the Association for the Advancement of Artificial Intelligence (AAAI).  He has been at Wisconsin since 1988, following the receipt of his PhD from the University of Illinois for his work on Explanation-Based Learning.  His current research interests include machine learning and computational biology, with an emphasis on using rich sources of training information, such as human-provided advice.  He served for three years as editor-in-chief of the AI Magazine and serves on the editorial board of about a dozen journals.  He chaired the 1998 International Conference on Machine Learning, co-chaired the First International Conference on Intelligent Systems for Molecular Biology in 1993, co-chaired the First International Conference on Knowledge Capture in 2001, was conference chair of the 2003 IEEE Conference on Data Mining, and co-chaired the 2007 International Conference on Inductive Logic Programming.  He was a founding member of both the board of the International Machine Learning Society and the board of the International Society for Computational Biology.  He co-edited, with Tom Dietterich, "Readings in Machine Learning."  His research has been supported by DARPA, NSF, NIH (NLM and NCI), ONR, DOE, AT&T, IBM, and NYNEX.