04 November 2009 @ 15:00 - 16:00

Seminar on "Probabilistic Graphical Networks: Toward a New Representation for Real-World Knowledge" by Dr. Leong Tze Yun, Department of Computer Science, School of Computing, National University of Singapore

Seminar Room, 8th Floor, Temasek Laboratories, 5A Engineering Drive 1, National University of Singapore

Abstract: A major challenge in computational cognition is in establishing a general knowledge representation that is adequately expressive to reflect the complexity and heterogeneity of real work knowledge, and sufficiently efficient to support reasoning, learning, and decision making processes based on such knowledge. Probabilistic graphical networks have recently emerged as a powerful mechanism that integrates subjective and objective information, and combines categorical concepts and uncertainty measures to support a wide range of cognitive tasks. In this talk, we examine the suitability of probabilistic graphical networks as a general representation for real-world knowledge. We motivate and explain the origins of the framework that evolved from well established knowledge representation formalisms from various disciplines, including philosophy, logic, cognitive science, probability theory, statistics, and decision theory. We introduce the formalism and the state-of-the-art development and application, and examine its strengths and weaknesses.

Speaker: Dr. Leong Tze-Yun is an Associate Professor of Computer Science and Vice Dean at the School of Computing, National University of Singapore. She directs the Medical Computing Laboratory in the School and leads the multi-disciplinary. Biomedical Decision Engineering Group at the University.

Dr. Leong received her S.B., S.M., and Ph.D. in Electrical Engineering and Computer Science from the Massachusetts Institute of Technology (MIT), USA. She has more than 20 years of professional and management background in biomedical informatics research and development. Her main expertise is in intelligent decision systems, and her research interests are in clinical and public health decision support, evidence-based medicine, data mining, and adaptive computing in complex and changing environments. Her current work focuses on integrating genotypic, phenotypic, and epidemiological information from heterogeneous sources to support biomedical decision making.

Dr. Leong has taught undergraduate and graduate courses both at MIT and NUS. With background in a broad spectrum of cognitive, knowledge-based and statistical data analysis techniques, she has led the design and implementation of various diagnostic, therapeutic, prognostic, and information integration systems in different domains, many of them in collaboration with leading hospitals in US and in Singapore. She regularly publishes in major scientific venues, and serves on the editorial boards and program committees for various international journals and conferences in artificial intelligence and biomedical computing. In particular, she is an Associate Editor of the International Journal of Artificial Intelligence in Medicine and Methods of Informatics in Medicine, and serves on the Editorial Board of the Journal of Biomedical Informatics. She also has business, training, and consultancy experiences in advanced decision support and data mining technologies for the healthcare, defense, and pharmaceutical industries. She is the founder and director of a high-tech company that specialises in computing technologies and services for next generation decision analytics, focusing on the healthcare industries.