Plenary SPEAKERS
Professor Petar M. Djuric
Professor Sun-Yuan Kung
Professor Erkki Oja

 

Professor Petar M. Djuric
Department of Electrical and Computer Engineering
Stony Brook University, USA
 
Homepage
Email: djuric@ece.sunysb.edu


Biography
Petar M. Djuric received his B.S. and M.S. degrees in electrical engineering from the University of Belgrade, in 1981 and 1986, respectively, and his Ph.D. degree in electrical engineering from the University of Rhode Island, in 1990. From 1981 to 1986 he was Research Associate with the Institute of Nuclear Sciences, Vinca, Belgrade. Since 1990 he has been with Stony Brook University, where he is Professor in the Department of Electrical and Computer Engineering. He works in the area of statistical signal processing, and his primary interests are in the theory of modeling, detection, estimation, and time series analysis and its application to a wide variety of disciplines including wireless communications and bio-medicine. Prof. Djuric has served on numerous technical committees for the IEEE and SPIE and has been invited to lecture at universities in the United States and overseas. He is the Area Editor of Special Issues of the Signal Processing Magazine and Associate Editor of the IEEE Transactions on Signal Processing. He is also Chair of the IEEE Signal Processing Society Committee on Signal Processing - Theory and Methods and is on the Editorial Board of Digital Signal Processing, the EURASIP Journal on Applied Signal Processing, and the EURASIP Journal on Wireless Communications and Networking. Prof. Djuric is a Member of the American Statistical Association and the International Society for Bayesian Analysis.

Lecture: Learning by particle filtering
Particle filtering is a methodology that has lately been attracting with increasing pace the attention of engineers and scientist. This methodology can be viewed as sequential learning of unknowns, where the unknowns may evolve with time or remain constant. Particle filtering is based on a random grid whose nodes are called particles and where the nodes have weights that account for their importance. The particles and the weights form a random measure, which approximates the distributions of the unknowns. In this presentation, first some of the basics of learning by particle filtering will be provided and then new developments and challenges in its theory and practice will be discussed.

 

Professor Sun-Yuan Kung
Department of Electrical Engineering
Princeton University
 
Homepage
Email: kung@ee.princeton.edu


Bibliography
Sun-Yuan Kung was born in Taiwan on January 2, 1950. He received the B.S. in Electrical Engineering from the National Taiwan University in 1971; M.S. in Electrical Engineering from the University of Rochester in 1974; and Ph.D. in Electrical Engineering from Stanford University in 1977.
From 1977 to 1987, he was on the faculty of Electrical Engineering-Systems at the University of Southern California. In 1984, he was a Visiting Professor at Stanford University and later in the same year, a visiting professor at the Delft University of Technology. Since September 1987, he has been a Professor in the Department of Electrical Engineering, Princeton University. He currently serves on the IEEE Technical Committees on VLSI Signal Processing and Neural Networks and an Editor-in-Chief of Journal of VLSI Signal Processing. Membership in Societies: IEEE (Fellow), ACM (Member).

Lecture: Data Mining and Visualization of DNA Microarray Data
Bioinformatics represents a natural convergence of life sciences, computer sciences, and device technologies. A gene expression or comparative genomic hybridization (CGH) experiment with DNA microarrays can simultaneously show information on thousands of genes, hence it represents one of the most common and promising instruments for genomic analysis. Understanding of gene function on a genomic scale can lead to many promising applications to drug design and disease classification. This talk will presents several machine learning approaches to two different types of microarray data mining problems: (1) gene clustering analysis a statistic problem and (2) chromosomal rearrangements detection a temporal problem. The gene clustering analysis is basically a statistic pattern recognition problem. The main challenge is to effectively handle and represent the usually huge amount of data set from microarrays. We shall present a hierarchical cluster discovery method for detection/validation of previously unrecognized tumors and a gene selection approach identifying the most critical gene subset responsible for the biological process that generates the patterns. On the other hand, the chromosomal rearrangement analysis must be formulated as a time series detection problem. Chromosomal rearrangements are one of primary causes of creation of new genetic material or alteration of genetic information. Undetected rearrangements can result in incorrect biological conclusion and ineffective medical treatments. We propose to adopt a novel and robust algorithm which is capable of high resolution breakpoint identification even for low SNR gene expression data. The proposed approach appears to be more robust than the traditional and correlate well with biological significance.

 

Professor Erkki Oja
Computer Science and Engineering
Helsinki University of Technology
 
Homepage
Email: Erkki.Oja@hut.fi


Biography
Erkki Oja is Director of the Neural Networks Research Centre and Professor of Computer Science at the Laboratory of Computer and Information Science, Helsinki University of Technology, Finland. He received his Dr.Sc. degree in 1977. He has been research associate at Brown University, Providence, RI, and visiting professor at Tokyo Institute of Technology. Dr. Oja is the author or coauthor of more than 260 articles and book chapters on pattern recognition, computer vision, and neural computing, and three books: "Subspace Methods of Pattern Recognition" (RSP and J.Wiley, 1983), which has been translated into Chinese and Japanese, "Kohonen Maps" (Elsevier, 1999), and "Independent Component Analysis" (J. Wiley, 2001). His research interests are in the study of principal component and independent component analysis, self-organization, statistical pattern recognition, and applying artificial neural networks to computer vision and signal processing. Dr. Oja is member of the editorial boards of several journals and has been in the program committees of several recent conferences including ICANN, IJCNN, and ICONIP. He is member of the Finnish Academy of Sciences, Fellow of the IEEE, Founding Fellow of the International Association of Pattern Recognition (IAPR), and President of the European Neural Network Society (ENNS).

Lecture: Beyond Independent Component Analysis
The linear blind source separation problem means solving a number of sources from their linear mixtures, when both the sources and the mixing coefficients are unknown. For instantaneous mixtures, a well-known technique is independent component analysis (ICA) which assumes that the sources are statistically independent. Many efficient algorithms for linear ICA exist. If constraints are added to the model, such as positivity of the sources and/or the mixing coefficients, solutions may still exist. If nonlinearities are introduced, the problem becomes much harder. For a general nonlinear mixing model, the problem is ill-defined and a unique solution does not exist. By suitably regularizing the problem, solutions can still be found. In the talk, some approaches to nonnegative ICA and nonlinear ICA are reviewed and illustrated by applications.