IEEE SMCals/06

Energy Conservation in Adaptive Filtering

Ali H. Sayed

Electrical Engineering Department

University of California at Los Angeles

Adaptive filters are systems that respond to variations in their environment by adapting their internal structure in order to meet certain performance specifications. Such systems are widely used in communications, biomedical applications, signal processing, and control. The performance of an adaptive filter is evaluated in terms of its transient behavior and its steady-state behavior. The former provides information about how fast a filter learns, while the latter provides information about how well a filter learns. Such performance analyses are usually challenging since adaptive filters are, by design, time-variant, nonlinear, and stochastic systems. For this reason, it has been common in the literature to study different adaptive schemes separately due to the differences that exist in their update equations.

The purpose of this talk is to provide an overview of an energy conservation approach to the performance analysis of adaptive filters. This framework is based on studying the energy flow through successive iterations of an adaptive filter and on establishing a fundamental energy conservation relation; the relation bears resemblance with Snell’s Law in optics and has far reaching consequences to the study of adaptive schemes. In this way, many new and old results can be pursued uniformly across different classes of algorithms.

In particular, the talk will highlight some recently discovered phenomena pertaining to the learning ability of adaptive filters. It will be seen that adaptive filters generally learn at a rate that is better than that predicted by least-mean-squares theory; that is, they are “smarter” than originally thought! It will also be seen that adaptive filters actually have two distinct rates of convergence; they learn at a slower rate initially and at a faster rate later; perhaps in a manner that mimics the human learning process.

Bio: Ali H. Sayed received his PhD in Electrical Engineering from Stanford University in 1992. He is Professor and Chairman of Electrical Engineering at UCLA where he also directs the Adaptive Systems Laboratory (www.ee.ucla.edu/asl). He has published widely in the areas of adaptive filtering, estimation theory, and signal processing for communications with over 250 articles and 4 books including the textbook Fundamentals a Adaptive Filtering (Wiley, NY, 2003). He is a Fellow of IEEE and serves as the Editor-in-Chief of the IEEE Transactions on Signal Processing. His research has received several recogniations including the 1996 IEEE D. G. Fink Prize, a 2002 Best Paper Award from the IEEE Signal Processing Society, the 2003 Kuwait Prize, the 2005 Frederick E. Terman Award, and two Best Student Paper Awards at international meetings (1999, 2001). He currently serves as a Distinguished Lecturer of the IEEE Signal Processing Society. He is also a member of the Publications and Award Boards of the IEEE Signal Processing Society, and serves a General Chairman of ICASSP 2008.

Learning in the Extreme: Lots of Data, Lots of Features, and/or Lots of Class Skew

Larry Hall

University of South Florida

Abstract: This talk covers applying data mining to data sets which may be considered extreme in some dimension. We include discussions of extremely large labeled data sets. Other types of extreme data sets may have few examples from an important class. For example, instances of disease tend to be greatly outweighed by normal patient data. It is important to focus on the small, important class(es) in building a model from such data. Another example of extreme data is, for example, MicroArray data. There may be many more features than there are examples. This requires a different type of feature selection.

Bio: Lawrence O. Hall is a Professor of Computer Science and Engineering at University of South Florida. He received his Ph.D. in Computer Science from Florida State University in 1986 and a B.S. in Applied Mathematics from the Florida Institute of Technology in 1980. He is a fellow of the IEEE. His research interests lie in distributed machine learning, data mining, pattern recognition and integrating AI into image processing. The exploitation of imprecision with the use of fuzzy logic in pattern recognition, AI and learning is a research theme. He has authored over 190 publications in journals, conference and books. Recent publications appear in Artificial Intelligence in Medicine, Neural Computation, Pattern Recognition Letters, JAIR, Journal of Machine Learning research, IEEE Transactions on Classifier Systems Workshop, and the FUZZ-IEEE conference. He co-edited the 2001 joint North American Fuzzy Information Processing Society (NAFIPS), IFSA conference proceedings. He was the co-Program Chair of NAFIPS 2004. He received the IEEE SMC Society Outstanidng contribution award in 2000. He received an Outstanding Research achievement award from the University of South Florida in 2004. A past president of NAFIPS. The former vice president for membership of the SMC society. He is the 2006 President of the SMC society. He was the Editor-in-Chief of the IEEE Transactions on Systems, Man and Cybernetics, Part B 2003 – 2005. Also, associate editor for IEEE Transactions on Fuzzy Systems, International Jounal of Intelligent Data Analysis, and International Journal of Approximate Reasoning.

Active noise and vibration control for periodic disturbances

Marc Bodson

Department of Electrical and Computer Engineering

University of Utah

Adaptive algorithms are used in active noise and vibration control problems to reject disturbances whose characteristics are unknown and vary with time. Often, the disturbances consist of one or more periodic signals, as they originate from nearby rotating machines. The talk will present an overview of methods available for the rejection of such disturbances in feedback systems. Various cases will be considered, including when the frequency of the disturbance is know and either fixed or time-varying, when the frequency is unknown and time-varying, and when the system in the feedback path in unknown and time-varying. Experimental results obtained on an active noise control testbed will also be presented.

Marc Bodson received a Ph.D. degree in Electrical Engineering and Computer Science from the University of California, Berkeley, in 1986. He obtained two M.S. degrees – one in Electrical Engineering and Computer Science and the other in Aeronautics and Astronautics – from the Massachusetts Institute of Technology, Cambridge MA, in 1982. In 1980, he received the degree of Ingénieur Civil Mécanicien et Electricien from the Université Libre de Bruxelles, Belgium. Currently, Marc Bodson is a Professor of Electrical & Computer Engineering at the University of Utah in Salt Lake City. He has been the Chair of the Department of Electrical & Computer Engineering since July 2003, and he was the Editor-in-Chief of IEEE Trans. on Control Systems Technology between January 2000 and December 2003. His research interests are in adaptive control, with application to electromechanical systems and aerospace. He is coauthor, with S. Sastry, of the book Adaptive Control: Stability, Convergence, and Robustness, published by Prentice-Hall in 1989 (the book is out-of-print but may be viewed as a PDF file at http://www.ece.utah.edu/~bodson/acscr).