IBM Research

NIPS 2006 Workshop

 


Call for Participation

NIPS 2006 Workshop
Revealing Hidden Elements of Dynamical Systems

Whistler, B.C - Canada
December 08 or 09, 2006
 http://www.haifa.il.ibm.com/Workshops/nips2006/

Background

Revealing and modeling the hidden state-space of dynamical systems is a fundamental problem in signal processing, control theory, and learning. Classical approaches to this problem include Hidden Markov Models, Reinforcement Learning, and various system identification algorithms. More recently, the problem has been approached by such modern machine learning techniques as kernel methods, Bayesian and Gaussian processes, latent variables, and the Information Bottleneck. Moreover, dynamic state-space learning is the key mechanism in the way organisms cope with complex stochastic environments such as biological adaptation. One familiar example of a complex dynamic system is the authorship system in the NIPS community. Such a system can be described by both internal variables, such as links between NIPS authors, and external environment variables, such as other research communities. This complex system, which generates a vast number of papers each year, can be modeled and investigated using various parametric and non-parametric methods.

In this workshop, we intend to review and confront different approaches to dynamical system learning, with various applications in machine learning and neuroscience. We plan to discuss relations between the different approaches, and address a range of questions and applications:
  • What are the special features of dynamical system learning that separate it from other learning problems?
  • What are the pros and cons of the current methods?
  • How can statistical and information theoretic techniques be combined with the theoretical structure of dynamical systems?
  • What kind of optimization principles for learning dynamics can be derived?
  • Are there generic features that can be extracted from time-series data?
  • How can we combine static and time series data for modeling dynamic systems?

In addition, we hope this workshop will familiarize the machine learning community with many real-world examples and applications of dynamical system learning. Such examples will also serve as the basis for the discussion of such systems in the workshop. A successful outcome of the workshop would be novel methods for learning and modeling from such data, as well as providing new conceptual frameworks for the general problem of adaptation to complex environments.

Tentative Speakers

  • Ziv Bar-Joseph, Center for Automatic Learning and Discovery (CALD) and the Computer Science Department, Carnegie Mellon University, USA (confirmed)
  • Jim Crutchfield, Computational Science and Engineering Center and Physics Department, University of California, Davis, USA (confirmed)
  • Irina Rish, IBM T.J. Watson Research Center (Hawthorne), USA (confirmed)
  • Pierre Baldi, School of Information and Computer Sciences, University of California Irvine, Irvine, USA (confirmed)
  • Naftali Tishby, School of Engineering and Computer Science, Hebrew University, Israel (confirmed)

Important Dates

Submission: October 31, 2006
Notification: November 13, 2006
NIPS 2006 Workshop: December 8 or 9, 2006


Workshop Organizers