| Monday, August 23 |
| 8:15-8:30 am |
Opening Remarks
Yu Hen Hu, Univ. of Wisconsin-Madison |
| 8:30-9:30 am |
Keynote Lecture:
Statistical Learning Theory and Induction Problem
Vladimir Vapnik
AT&T Labs Research, USA
Chair: Klaus-Robert Müller, GMD-FIRST |
| 9:30-9:40 am |
Coffee Break |
| 9:40-12:00 am |
Special Session: Support Vector Machines
Chair: Klaus-Robert Müller, GMD-FIRST
|
- SUPANOVA - A Sparse Transparent Modelling Approach
Steve R. Gunn, University of Southampton, UK
Martin Brown , Unilever, UK
- Sequential Support Vector Machines
N. de Freitas, M. Milo, P. Clarkson, M. Niranjan, and A. Gee, Cambridge University, UK
- Fisher Discriminant Analysis with Kernels
Sebastian Mika, Gunnar Rätsch, Jason Weston, Bernhard Schölkopf and Klaus-Robert Müller, GMD First, Germany
- The Application of Support Vector Machines with Gaussian Kernels for Overcoming Co-Channel Interference
Felix Albu, Falculty of Elect. and Telecom., Romania
Dominique Martinez, CNRS-LAAS, France
- A Gaussianity Measures for Blind Source Separation Insensitive to the Sign of Kurtosis
Hsiao-Chun Wu and Jose Principe, Univ. of Florida, USA
- Robust Machine Fault detection with Independent Component Analysis and Support Vector Data Description
Alexander Ypma, David M. J. Tax and Robert P. W. Duin, Delft University of Technology, Netherlands
- View-based 3D Object Recognition with Support Vector Machines
Danny Roobaert, Royal Institute of Technology, Sweden, K.U. Leuven, Belgium
Marc M. Van Hulle, K.U. Leuven, Belgium
|
|
| 12:00 am-1:00 pm |
Lunch |
| 1:00-2:00 pm |
Keynote Lecture:
Margin-Like Quantities and Generalized Approximate Cross Validation for Support Vector Machines
Grace Wahba
Department of Statistics
University of Wisconsin-Madison, USA
Chair: Jan Larsen, Techn. Univ. of Denmark
|
| 2:00-3:00 pm |
Algorithm and Architectures 1
Chair: David Miller, Pennsylvania State University
|
- Gradient Based Adaptive Regularization
Robert Eigenmann, Josef A. Nossek, Munich University of Technology, Germany
- Global Stability in Delayed Celluar Neural Networks
Jinde Cao, Yunnan University, China
- Weight Estimation for the learning of Modular Perceptron Networks
Y.P. Lee, Hsin-Chia Fu, National Chiao Tung University, Taiwan, ROC
|
|
| 3:00-3:20 pm |
Coffee Break |
| 3:20-4:00 pm |
Algorithm and Architectures 1 (cont.)
Chair: David Miller, Pennsylvania State University
|
- General Statistical Inference by an Approximate Application of the Maximum Entropy Principle
Lian Yan, David J. Miller, Pennsylvania State University, USA
- On Training Piecewise-Linear Networks
A. Atiya, Caltech, USA
E. Gad, Carleton Univ., USA
S. Shaheen, Cairo Univ., Egypt
A. El-Dessouky, ERI, Egypt
|
|
| 4:00-6:00 pm |
Poster Session 1: Support Vector Machines, Algorithm and Architectures
Chair: Scott Douglas, Southern Methodist University
|
Using Class-Center Vectors to Build Support Vector Machines
Xuegong Zhang, Tsinghua University, China
A Gaussianity Measures for Blind Source Separation Insensitive to the Sign of Kurtosis
Hsiao-Chun Wu and Jose Principe, Univ. of Florida, USA
Robust Machine Fault Detection with Independent Component Analysis and Support Vector Data Description
Alexander Ypma, David M. J. Tax and Robert P. W. Duin, Delft University of Technology, Netherlands
View-based 3D object Recognition with Support Vector Machines
Danny Roobaert, Royal Institute of Technology, Sewden, K.U. Leuven, Belgium
Marc M. Van Hulle, K.U. Leuven, Belgium
Hidden Markov Mixtures of Experts for Prediction of Non-Stationary Dynamics
S. Liehr, K. Pawelzik, Univ. of Bremen, Germany
J. Kohlmorgen, S. Lemm, K.-R. Müller, GMD First, Germany
Nearest-Prototype Classifier Design by Deterministic Annealing with Random Class Labels
Ertem Tuncel, Kenneth Rose, Univ. of California, Santa Barbara, USA
A Comparison Among Feature Selection Methods Based on Trained Networks
Mercedes Fernandez, Carlos Hernandez, Jaume-I University, Spain
Robust Neural Network Online-Learning in Time-Variant Regression Models
Thomas Briegel, Volker Tresp, Siemens AG, Munich, Germany
Separation of an Overlapped Signal using Speech Production Models
Hideyuki Watanabe, Satoru Fujita, ATR, Japan
Shigeru Katagiri, ATR and NTT Communication Science Laboratories, Japan
Geometrical Structures of FIR Manifold and their Application to Multichannel Blind Deconvolution
L. -Q. Zhang, A. Cichocki, and S. Amari, RIKEN, Japan
|
|
| 6:00-7:30 pm |
Dinner (on your own) |
| 7:30-9:00 pm |
Welcome Reception |
| Tuesday, August 24 |
| 8:00-9:00 am |
Keynote Lecture: Using Neuronets in Multimodal Human Computer Interfaces
Thomas S. Huang
Beckman Institute for Advanced Science and Technology
University of Illinois at Urbana-Champaign
Chair: Yu Hen Hu, Univ. of Wisconsin-Madison
|
| 9:00-9:40 am |
Applications 1: Intelligent Human Computer Interfaces
Chair: Yu Hen Hu, Univ. of Wisconsin-Madison
|
- Statistical Multimodal Integration for Intelligent HCI
Lizhong Wu, Sharon Oviatt, and Philip Cohen, Oregon Graduate Institute, USA
- Hierarchy of Probabilistic Principal Component Subspaces for Data Mining
Lan Luo, Yue Wang, The Catholic University of America, Washington DC
San Yun Kung, Princeton University, Princeton, USA
|
|
| 9:40-10:00 am |
Coffee Break |
| 10:00-12:00 am |
Algorithm and Architectures 2
Chair: Amir Assadi, Univ. of Wisconsin-Madison
|
- A Control Theory Formulation for Random Variate Generation
M. Magdon-Ismail and A. Atiya, Caltech, USA
- A Spiking Neural Network Architecture for Nonlinear Function Approximation
Nicolangelo Iannella and Andrew Back, RIKEN, Japan
- Partial Likelihood Methods for Probability Density Estimation
Hongmei Ni, Tülay Adali, and Bo Wang, University of Maryland, Baltimore County, USA
- Why a Nonlinear Solution for a Linear Problem?
Tülay Adali, Univ. of Maryland, Baltimore County, USA
- Approximation by Random Networks with Bounded Number of Layers
Erol Gelenbe, Univ. of Central Florida, USA
Z.-H. Mao, Y.-D. Li, Tsinghua Univ., China
- Natural Power Method for Fast Subspace Tracking
Yingbo Hua, Young Xiang, Tiangping Chen, Karim Abed-Meraim, and
Yongfeng Miao, Univ. of Melbourne, Australia
|
|
| 12:00 am-1:00 pm |
Lunch |
| 1:00-3:20 pm |
Algorithm and Architectures 2 (cont.)
Chair: Amir Assadi, Univ. of Wisconsin-Madison
|
- Robust Neural Network Online-Learning in Time-Variant Regression Models
Thomas Briegel, Volker Tresp, Siemens AG, Munich, Germany
- A Comparison Among Feature Selection Methods Based on Trained Networks
Mercedes Fernandez and Carlos Hernandez, Jaume-I University, Spain
- Sizing of the Multilayer Perceptron via Modular Networks
Hema Chandrasekaran, University of Texas at Arlington, USA
Kyung K. Kim, Lockheed Martin Tactical Aircraft Systems, USA
Michael T. Manry, University of Texas at Arlington, USA
- On Optimal Data Split for Generalization Estimation and Model Selection
Jan Larsen and Cyril Goutte, Techn. Univ. of Denmark, Denmark
- Time Series Classification using Adaptive Dynamic Targets
Ernst Haselsteiner, Technical University Graz, Austria
- Some Analytical Results on Critic-driven Ensemble Classification
David J. Miller and Lian Yan, Pennsylvania State University, USA
- Neural Architectures for Parametric Estimation of a Posteriori Probabilities by Constrained Conditional Density Functions
Juan Ignacio Arribas, Jesus Cid-Sueiro, University of Valladolid, Spain
Tülay Adali, University of Maryland, Baltimore County, USA
Anibal R. Figueiras-Vidal, University Carlos III, Spain
|
|
| 3:20-3:40 pm |
Coffee Break |
| 3:40-6:00 pm |
Poster Session 2: Applications
Chair: Kevin R. Farrell, T-NETIX, Inc.
|
Neuro Bayesian Blind Equalization with BER Estimation in Digital Channels
L. M. San Jose Revuelta, J. Cid Sueiro, Univ. de Valladolid, Spain
Discrimination of Cylinders with Different Wall Thicknesses using Neural Networks and Simulated Dolphin Sonar Signals
Lars Nonboe Andersen, Techn. Univ. of Denmark, Denmark
Whitlow Au, University of Hawaii, USA
Jan Larsen and Lars Kai Hansen, Techn. Univ. of Denmark, Denmark
Adaptive Web Caching Using Logistic Regression
Annie P. Foong, Yu-Hen Hu, and Dennis M. Heisey, University of Wisconsin-Madison, USA
Near-Optimal Flight Load Synthesis Using Neural Nets
Michael T. Manry, Cheng-Hsiung Hsieh, and Hema Chandrasekaran, University of Texas at Arlington, USA
Two-Dimensional Joint Process Adaptive Filtering via Principal Component Support Region
Dai I. Kim and P. De Wilde, Imperial College, Univeristy of London, UK
Clustering Approach to Square and Non-square Blind Source Separation
Marc M. Van Hulle, K.U. Leuven, Belgium
Learning Gestalt Of Surfaces In Natural Scenes
Amir Assadi, UW-Madison, USA
Stephen Palmer, UC-Berkeley, USA
Hamid Eghbalnia, UW-Madison, USA
An Adaptive Metric Learning Procedure for Reconfigurable Facial Signature Authentication
Takami Satonaka, Tatsuo Otsuki, and Takao Chikamura, Matsushita Electronics Corpration, Osaka, Japan
A Topology Independent Active Contour Tracking
Dongxiang Xu, Jenq-Neng Hwang, Univ. of Washington, USA
Hierarchy of Probabilistic Principal Component Subspaces for Data Mining
Lan Luo, Yue Wang, The Catholic University of America, Washington DC
San Yun Kung, Princeton University, Princeton, USA
|
|
| 7:30 pm- |
Galla Banquet |
| Wednesday, August 25 |
| 8:00-10:00 am |
Applications 2: Blind Source Separation
Chair: Marc Van Hulle, K.U. Leuven
|
- All-Pass vs. Unit-Norm Constraints in Contrast-Based Blind Deconvolution
Scott C. Douglas, Southern Methodist University, USA
S.-Y. Kung, Princeton University, USA
- A Stable and Robust ICA Algorithm Based on t-Distribution and Generalized Gaussian Distribution Model
Jianting Cao, Sophia University, Tokyo and RIKEN Brain Science Institute, Japan
Noboru Murata, RIKEN Brain Science Institute, Japan
- Separation of an Overlapped Signal using Speech Production Models
Hideyuki Watanabe, Satoru Fujita, ATR, Japan
Shigeru Katagiri, ATR and NTT Communication Science Laboratories
- Geometrical Structures of FIR Manifold and their Application to Multichannel Blind Deconvolution
L. -Q. Zhang, A. Cichocki and S. Amari, Brain Science Institute, RIKEN, Japan
- Clustering Approach to Square and Non-square Blind Source Separation
Marc M. Van Hulle, K.U. Leuven, Belgium
- Separation of Acoustic Signals using Self-Organizing Neural Networks
Temujin Gautama and Marc Van Hulle, K.U. Leuven, Belgium
|
|
| 10:00-10:20 am |
Coffee Break |
| 10:20-11:00 am |
Applications 2: Blind Source Separation (cont.)
Chair: Marc Van Hulle, K.U. Leuven
|
- Maximum Likelihood Blind Source Separation in Gaussian Noise
Joaquín Míguez and Luis Castedo, Univ. da Coruña, Spain
- A Suboptimal Bayesian Equalizer using an Nonlinear Multilayer Combiner
Sooyong Choi and Daesik Hong, Yonsei University, Seoul, Korea
|
|
| 11:00-12:00 am |
Applications 3
Chair: Naonori Ueda, NTT Communication Science Labs
|
- A New Tool for the Market Research using a Modified Auto-Associative Memory
V. Giménez- Martínez, J. Castellanos, and F. de Mingo, Univ. Politécnica de Madrid, Spain
- Pattern Classification using a Mixture of Factor Analyzers
Naonori Ueda, Ryohei Nakano, NTT Communication Science Labs, Kyoto, Japan
Zoubin Ghahramani, Geoffery Hinton, University College London, UK
- On Condition Monitoring of Exhaust Valves in Marine Diesel Engines
T. L. Fog, MAN B&W Diesel A/S, Copenhagen, Denmark
L. K. Hansen, J. Larsen, Techn. Univ. of Denmark, Denmark
H. S. Hansen, L. B. Madsen, P. Sørensen, E.R. Hansen, and P.S. Pedersen, MAN B&W Diesel A/S, Copenhagen, Denmark
|
|
| 12:00 am-1:00 pm |
Lunch |
| 1:00-3:00 pm |
Applications 4: Image Processing
Chair: Jenq-Neng Hwang, Univ. of Washington
|
- Applications of SOAR to Monochromatic Image Restoration
Yusuf Ozturk and Huseyin Abut San Diego State University, USA
- Efficient Approximation of a Neural Filter for Quantum Noise Removal in X-Ray Images
Kenji Suzuki, Isao Horiba, Aichi Prefectural University, Aichi, Japan
Noboru Sugie, Meijo University, Nagoya, Japan
- Learning Gestalt Of Surfaces In Natural Scenes
Amir Assadi, UW-Madison, USA
Stephen Palmer, UC-Berkeley, USA
Hamid Eghbalnia, UW-Madison, USA
- Rate-Constrained Self-Organising Neural Maps and Efficient Psychovisual Methods for Low Bit Rate Video Coding
Keith Ferguson and Nigel Allinson, UMIST, Manchester, UK
- Selective Image Smoothing via Dyadic Wavelet-based Conduction Equation
Chwen-Jye Sze, Hong-Yaun Mark Liao, Shih-Kun Huang and Chun-Shien Lu, Academia Sinica, Taiwan, ROC
- An Adaptive Metric Learning Procedure for Reconfigurable Facial Signature Authentication
Takami Satonaka, Tatsuo Otsuki, and Takao Chikamura, Matsushita Electronics Corpration, Osaka, Japan
|
|
| 3:00-3:20 pm |
Coffee Break |
| 3:20-4:00 pm |
Applications 4: Image Processing (cont.)
Chair: Jenq-Neng Hwang, Univ. of Washington
|
- Global Color Image Segmentation Strategies: Euclidian Distance vs. Vector Angle
Slawo Wesolkowski, NCR Canada Ltd., Waterloo and Univ. of Waterloo, Canada
Robert D. Dony, Univ. of Guelph, Guelph, Canada
M. E. Jernigan, Univ. of Waterloo, Canada
- A Topology Independent Active Contour Tracking
Dongxiang Xu and Jenq-Neng Hwang, Univ. of Washington, USA
|
|
| 4:00-5:20 pm |
Applications 5
Chair: Hervé Bourlard, EPFL
|
- Model Combination and Weight Selection Criteria for Speaker Verification
Kevin R. Farrell, T-NETIX/SpeakEZ Inc., Englewood, CO, USA
- Subspace Techniques in Speech Enhancement
Gavin Smith, Cambridge University, Cambridge, UK
Mahesan Niranjan, T. Robinson, Sheffield University, Sheffield, UK
- Multi-Channel Piecewise Selective Averaging of Cognitive Evoked Potentials with Variable Latency
Arthur Flexer, The Austrian Research Institute for Artificial Intelligence, Vienna, Austria
Herbert Bauer, Univ. of Vienna, Austria
- A Comparative Study of a Hidden Markov Model Detector for Atrial Fibrillation
Brian Young, Done Brodnick, and Randy Spaulding, GE Marquette Medical Systems, USA
|
|
| 5:20-5:40 pm |
Wrap Up and Farewell |
Download advance program in PDF format
