CHIN%PLU@IO.ARC.NASA.GOV (05/07/88)
*************************************************************************** National Aeronautics and Space Administration Ames Research Center SEMINAR ANNOUNCEMENT SPEAKER: Smadar Kedar-Cabelli Rutger University TOPIC: Formulating Concepts and Analogies According to Purpose ABSTRACT: This talk describes research within the {\it explanation-based generalization} (EBG) framework, a framework for producing deductive generalizations from single examples. Despite recent progress, EBG methods exhibit an important limitation: they are incapable of determining which concepts are useful ones to acquire. More robust generalizers must be able to automatically determine which concepts to acquire based on the {\it purpose} of the learning, since concepts acquired for one purpose may not be appropriate for another. Our notion of the purpose of the learning is to acquire concepts which will benefit an associated performance system. Two open issues become apparent once EBG is associated with a performance system: How can EBG acquire target concepts and definitions appropriate for the performance system? Further, could the acquired target concept definitions be used to improve subsequent performance? The research focuses and investigates these issues in the context of a specific type of performance system -- a state-space planner. The approach is to provide EBG with explicit knowledge of the planner and specific planning task. The {\it purposive concept formulation} and {\it purposive explanation replay} methods, respectively, provide solutions to the open problems. We describe the prototype systems (PurForm and REPeat) which provide experimental support for these methods. The results confirm that a learning system {\it can} formulate concepts and analogies sensitive to the purpose of the learning in restricted planning situations. We discuss further extensions suggested by these results. BIOGRAPHY: Smadar Kedar-Cabelli has recently received her Ph.D. from the Department of Computer Science at Rutgers University. She is currently a research assistant at Rutgers, and a consultant for the Learning Systems Group at Siemens Research and Technology Laboratories in Princeton. Her research in machine learning focuses on open problems within the explanation-based generalization (EBG) framework. She has published a number of recent papers describing the dissertation results. A paper presented at the 1987 National Conference for Artificial Intelligence describes results on formulating concept according to purpose. A paper describing the close relationship of EBG and resolution-theorem proving was presented at the Fourth International Machine Learning Workshop in 1987. Earlier papers include a journal paper in Machine Learning, published jointly with Mitchell and Keller, introducing the explanation-based generalization framework. --------------------- DATE: Thursday TIME: 2:00 - 3:00 pm BLDG. 244 Room 209 May 19, 1988 -------------- POINT OF CONTACT: Marlene Chin PHONE NUMBER: (415) 694-6525 NET ADDRESS: chin%plu@ames-io.arpa *************************************************************************** VISITORS ARE WELCOME: Register and obtain vehicle pass at Ames Visitor Reception Building (N-253) or the Security Station near Gate 18. Do not use the Navy Main Gate. Non-citizens (except Permanent Residents) must have prior approval from the Director's Office one week in advance. Submit requests to the point of contact indicated above. Non-citizens must register at the Visitor Reception Building. Permanent Residents are required to show Alien Registration Card at the time of registration. ***************************************************************************