program.html

---
title: Program
layout: default
---

<h2>Program for AISTATS 2001</h2>
<br>
<p>
<li><h3><u>    Thursday January 4: tutorials</u>
</h3>
<p>
     10.00 -- 12.00
<br>
     David Edwards (Novo Pharmaceutical Company)
<br>
     "Graphical models for mixed data : an introduction to the theory and
     practice"
<br>
<p>
     2.00  --  4.00
<br>
     Yoav Freund    (AT&T Research)
<br>
     "Predicting More -- Assuming Less, new methodologies in statistical
     inference"
<br>
<p>
     Coffee Break     
<br>
<p>
     4.30  --  6.30
<br>
     David MacKay  (University of Cambridge)
<br>
     "Error correcting codes and belief propagation"
<br>
<p>
<p>
<li><h3><u>    Friday January 5 </u>
</h3>
<p>
     7.30  --  8.45  Breakfast
<br>
<p>
     8.45  --  9.00  Opening
<br>
<p>
     Morning 1: Clustering 
<br>
<p>
     9.00  --  9.30     
<br>
     The Learning Curve Method Applied to Clustering
<br>
     Chris Meek, Bo Thiesson, David Heckerman
<br>
<p>
     9.30  -- 10.00
<br>
     Bootstrapping Self-Organizing Maps to Assess the Statistical
     Significance of Local Proximity
<br>
     Eric de Bodt, Marie Cottrell
<br>
<p>
     10.00 -- 10.30
<br>
     A Random Walks View of Spectral Segmentation
<br>
     Marina Meila, Jianbo Shi
<br>
<p>
Coffee break 
<br>
<p>
     Morning 2: Speech/language
<br>
<p>
     11.00 -- 11.30
<br>
     Distributional Similarity, Frequency, and the Skew Divergence
<br>
     Lillian Lee
<br>
<p>
     11.30 -- 12.00
<br>
     Handling Missing and Unreliable Information in Speech Recognition
<br>
     Phil Green, Martin Cooke, Jon Barker, Ljubomir Josifovski
<br>
<p>
     12.00 --  2.00 Conference lunch
<br>
<p>
<p>
     Afternoon: Graphical Model search
<br>
<p>
     2.00  --  2.30
<br>
     Finding an optimal chain is harder than finding an optimal tree
<br>
     Chris Meek
<br>
<p>
     2.30  --  3.00
<br>
     An Anytime Algorithm for Causal Inference
<br>
     Peter Spirtes
<br>
<p>
     3.00  --  3.30
<br>
     A Simulation Study of Three Related Causal Data Mining Algorithms
<br>
     Subramani Mani, Gregory Cooper
<br>
<p>
     Coffee break 
<br>
<p>
     4.00  --  5.00 Poster preview
<br>
<p>
     Dinner (not provided)
<br>
<p>
     7.30  -- 10.00 Poster session 1
<br>
<p>
<p>
<li><h3> <u>   Saturday January 6 </u>
</h3>
<p>
     7.30  --  9.00  Breakfast
<br>
<p>
     Morning 1: Boosting
<br>
<p>
     9.00  --  9.30
<br>
     Boosting for Regression and Classification: Some Views from Analogy
<br>
     Wenxin Jiang
<br>
<p>
     9.30  -- 10.00
<br>
     Online Bagging and Boosting
<br>
     Nikunj Oza, Stuart Russell
<br>
<p>
     Coffee break 
<br>
<p>
     Morning 2: Kernel methods
<br>
<p>
     10.30 -- 11.00
<br>
     An improved training algorithm for kernel Fisher discriminants
<br>
     Sebastian Mika, Alexander Smola, Bernhard Schoelkopf
<br>
<p>
     11.00 -- 11.30
<br>
     A Kernel Approach for Vector Quantization with Guaranteed Distortion
     Bounds
<br>
     Michael Tipping, Bernhard Schoelkopf
<br>
<p>
     11.30 -- 12.30 Poster preview
<br>
<p>
     Lunch (not provided)
<br>
<p>
     5.00  --  7.30 Poster session 2
<br>
<p>
     7.30  -- 10.00 Conference banquet 
<br>
<p>
<p>
<li><h3> <u>   Sunday, January 7 </u>
</h3>
<p>
     7.30  --  9.00 Breakfast 
<br>
<p>
     Morning 1: Dynamic Bayesian networks
<br>
<p>
     9.00  --  9.30
<br>
     Products of Hidden Markov Models
<br>
     Andrew Brown, Geoffrey Hinton
<br>
<p>
     9.30  -- 10.00
<br>
     Solving Hidden-Mode Markov Decision Problems
<br>
     Samuel Choi, Nevin Zhang, Dit-Yan Yeung
<br>
<p>
     10.00 -- 10.30
<br>
     Can the Computer Learn to Play Music Expressively?
<br>
     Christopher Raphael
<br>
<p>
     Coffee break
<br>
<p>
     Morning 2: Multiple Models
<br>
<p>
     11.00 -- 11.30
<br>
     Hyperparameters for Soft Bayesian Model Selection
<br>
     Adrian Corduneanu, Christopher Bishop
<br>
<p>
     11.30 -- 12.00
<br>
     Managing Multiple Models
<br>
     Hugh Chipman, Edward George, Robert McCulloch
<br>
<p>
     12.00 -- 12.15 Closing
<br>
<p>
<hr>
<p>
<h3><u>Papers in poster session 1 (Friday January 5) </u></h3>
<p>
Image Decomposition and Tracking using Dynamic Positional Trees
<br>
Amos Storkey, Christopher Williams
<br>
<p>
Profile Likelihood in Directed Graphical Models from BUGS Output
<br>
Malene Hojbjerre
<br>
<p>
Stochastic System Monitoring and Control
<br>
Gregory Provan
<br>
<p>
Comparing Prequential Model Selection Criteria in Supervised Learning of
Mixture Models
<br>
Petri Myllymaki, Petri Kontkanen, Henry Tirri
<br>
<p>
Variational Learning for Multi-Layer Networks of Linear Threshold Units
<br>
Neil Lawrence
<br>
<p>
Models for Conditional Probability Tables in Educational Assessment
<br>
Russell Almond, DiBello Lou, Frank Jenkins, Robert Mislevy, Deniz Senturk,
<br>
Linda Steinberg, Duanli Yan
<br>
<p>
Discriminant Analysis on Dissimilarity Data : a New Fast Gaussian like
Algorithm
<br>
Guerin-Dugue Anne, Celeux Gilles
<br>
<p>
Bayesian Support Vector Regression
<br>
Martin Law, James Kwok
<br>
<p>
Message Length as an Effective Ockham's Razor in Decision Tree Induction
<br>
Scott Needham, David Dowe
<br>
<p>
Dual perturb and combine algorithm
<br>
Pierre Geurts
<br>
<p>
Bagging and the Bayesian Bootstrap
<br>
Merlise Clyde, Herbert Lee
<br>
<p>
On the correspondence between partially-observable Markov decision processes
and Bayes-adaptive Markov decision processes
<br>
Michael Duff
<br>
<p>
Learning mixtures of smooth nonuniform deformation fields for probabilistic
image matching
<br>
Nebojsa Jojic, Brendan Frey, Patrice Simard, David Heckerman
<br>
<p>
Clustering in high dimensions: modular mixture models
<br>
Hagai Attias
<br>
<p>
<hr>
<p>
<h3><u>Papers in poster session 2 (Saturday January 6)</u></h3>
<p>
Learning in Bayesian networks with mixed variables
<br>
Susanne Bottcher
<br>
<p>
Instrumental Variable Estimation of Causal Influence without Linearity
<br>
Richard Scheines, Greg Cooper
<br>
<p>
On Parameter Priors for Discrete DAG Models
<br>
Dan Geiger, Dmitry Rusakov
<br>
<p>
The Efficient Propagation of Arbitrary Subsets of Beliefs in Discrete-Valued
Bayesian Networks
<br>
Duncan Smith
<br>
<p>
Another look at sensitivity of Bayesian networks to imprecise probabilities
<br>
Oscar Kipersztok, Hai-Qin Wang
<br>
<p>
Some variations on variation independence.
<br>
Philip Dawid
<br>
<p>
On searching for optimal classifiers among Bayesian networks 
<br>
Robert Cowell
<br>
<p>
Geographical clustering of cancer incidence by means of Bayesian networks
and conditional Gaussian networks
<br>
J. M. Pena, I. Izarzugaza, J. A. Lozano, E. Aldasoro and P. Larranaga
<br>
<p>
Statistical Aspects of Stochastic Logic Programs
<br>
James Cussens
<br>
<p>
Temporal Matching under Uncertainty
<br>
Ahmed Tawfik, Greg Scott
<br>
<p>
Using Unsupervised Learning to Guide Resampling in Imbalanced Data Sets
<br>
Adam Nickerson, Nathalie Japkowicz, Evangelos Milios
<br>
<p>
Information-Theoretic Advisors in Invisible Chess
<br>
Ariel Bud, Ingrid Zukerman, David Albrecht, Ann Nicholson
<br>
<p>
A Non-Parametric EM-Style Algorithm for Filling-In Missing Values
<br>
Rich Caruana
<br>
<p>
Predicting with Variables Constructed from Univariate Temporal Sequences
<br>
Mehmet Kayaalp, Greg Cooper, Gilles Clermont