MSRI Workshop on Nonlinear Estimation and Classification
Berkeley, California
March 19-29, 2001

Invited Program

Below we list a number of rough, proposed sessions. Those speakers who have agreed to participate (at least tentatively subject to scheduling constraints) appear in the sessions below. Those who have accepted the invitation from MSRI and will definitely participate appear in bold face.

These sessions will undoubtedly change when we receive abstracts from each speaker. At that point, we anticipate reshuffling the sessions to better focus the separate ideas.

Week 1 Week 2
Monday Monday
Nonlinear classification (discussion)
  1. Gauss Mixture Vector Quatization: Clustering Gauss Mixtures with the Minimum Discrimination Information Distortion for Modeling, Compression, and Classification
    Robert Gray (Stanford, EE)
  2. Some experiments with ensemble methods for classification and conditional density estimation
    Tom Dietterich (Oregon State, CS)
  3. Dilution priors for model uncertainty
    Edward I. George (UT, Stat)
Approximation theory and signal processing (discussion)
  1. Richard Baraniuk (Rice, EECS)
  2. Martin Vetterli (EPFL,EECS)
  3. Harmonic analysis of BV
    Ronald DeVore (USC, Math)
Tuesday Tuesday
SVMs and functional data analysis
  1. Optimal properties and adaptive tuning of support vector machines (SVM's)
    Grace Wahba (Wisconsin-Madison, Stat)
  2. Detecting a menstrual cycle signal in flourescence spectroscopy of the cervix
    Dennis Cox (Rice, Stat)
Planned Outing
Modeling of complex objects
  1. Coarse-to-Fine Classification and Visual Indexing
    Donald Geman (UMass, Math)
  2. From Bits to Information: Theory and Applications of Learning Machines
    Tomaso Poggio (MIT, Cog Sciences)
  3. Gradient-based learning in heterogeneous structures
    Yann LeCun (AT&T, Speech and Image Processing)
Poster Session #2
Wednesday Wednesday
Theory
  1. Extended linear modeling with splines
    Charles Stone (Berkeley)
  2. Bayesian approaches in nonparemetric estimation problems
    Linda Zhao (UPenn)
  3. Information theoretic bounds for mixture modeling and model selection
    Andrew Barron (Yale)
Poster Session #1
Case studies
  1. Environmental monitoring using a time series of satellite images and other spatial data sets
    Harri Kiiveri (CSIRO, Perth, Australia)
  2. Bayesian multidimensional scaling and choice of dimension
    Adrian Raftery (UW, Stat)
Planned Outing
Thursday Thursday
Modeling complex systems
  1. Variational inference for clustering and classification
    Michael Jordan (Berkeley, CS and Stat)
  2. Modeling internet traffic
    Bill Cleveland (Bell Labs, Stat)
  3. A simple model for a complex system: Predicting travel times on freeways
    John Rice (Berkeley, Stat)
Machine learning and optimization (discussion)
  1. Jerome Friedman (Stanford, Stat)
  2. Logistic Regression, AdaBoost and Bregman Distances
    Rob Schapire (AT&T Labs, CS)
  3. Where does mainstream optimization fit?
    Margaret H. Wright (Bell Labs, Math)
  4. Leo Breiman (Berkeley, Stat)
Friday Friday
Future Directions
  1. Nonlinear estimation and classification: challenges from genomics.
    Terry Speed(Stat, UCB)
  2. The complex statistics of images
    David Mumford (Brown, Applied Math)
  3. Interactions between data analysis of natural images, biological vision and mathematical analysis
    David Donoho (Stanford, Stat)