CSNA/97 CONFERENCE PROGRAM
American University
Mary Graydon Center/ Butler Pavilion
Washington DC
June 12-15, 1997
June 12: Short Courses, Mixer, and Board Meeting
8:30- 5:00 INTRODUCTION TO CLASSIFICATION AND CLUSTERING
Stephen C. Hirtle, University of Pittsburgh
Pierre Legendre, Universite de Montreal
Glenn W. Milligan, Ohio State University
8:30- 9:00 Registration
9:00-10:30 Session I: Exploratory Data Analysis: An Overview.
Stephen C. Hirtle, University of Pittsburgh
10:30-10:45 Coffee
10:45-12:45 Session II: Cluster Analysis.
Pierre Legendre
12:45- 1:45 Lunch
1:45- 3:45 Session III: Clustering Validation and Comparison.
Glenn W. Milligan, Ohio State University
3:45- 4:00 Coffee
4:00- 5:00 Session IV: Example, Analysis, Discussion.
Milligan, Hirtle, Legendre
8:30- 5:00 MULTIVARIATE NONPARAMETRIC REGRESSION
David L. Banks, NIST
Chris Genovese, Carnegie Mellon University
Stephanie Land, Carnegie Mellon University
8:30- 9:00 Registration
9:00- 9:50 Curse of Dimensionality; Model Assessment; Inference
10:00-10:50 Smoothing; Variable Selection
11:00-11:50 Generalized Additive Models; Backfitting
12:00- 1:00 Lunch
1:00- 1:50 Projection Pursuit; Neural Networks; Barron's Theorem
2:00- 2:50 RPR, MARS, ACE, AVAS, Loess
3:00- 3:50 Wavelets
4:00- 5:00 Guidelines for Applications
6:30 Registration and Reception
7:30 Board Meeting
June 13: Conference Sessions and Banquet
8:45- 9:15 OFFICIAL WELCOME.
Olga Cordero-Bra~na, Conference Organizer
Lynn Stallings, Chair of Dept. of Math/Stat/Computer Sci.
9:15-10:15 Plenary I: NONPARAMETRIC DISCRIMINANT ANALYSIS
Organizer: David Banks, NIST
Flexible Discriminant and Mixture Models
Rob Tibshirani, University of Toronto
10:15-10:30 Coffee.
10:30-11:30 Plenary II: BLOCK CLUSTERING
Organizer: Olga Cordero-Bra~na, American University
New Ideas in Block Clustering
John Hartigan, Yale University
11:30-11:45 Program announcements.
11:45- 1:00 Lunch.
1:00- 2:45 Invited Session 1: PUBLIC HEALTH STATISTICS
Organizer: Demissie Alemayehu, Columbia University
and Pfizer, Inc.
Adaptive Signal Regression
Stephanie Land, Carnegie Mellon University
Application of Weighted Likelihood Equations in the
Classification of Laboratory Data.
Marianthi Markatou, Columbia University
Establishing Normal Reference Ranges by Regression
Quantiles
Xuming He, University of Illinois
Discussant: Herman Friedman, New York
Contributed Session 1: PROBLEMS WITH TREES
Chair: Pascale Rousseau, Universite du Quebec a Montreal
Bootstrap Significance Levels for Phylogenetic Trees
Generated from Correlated Characters
Waleed S. AlGharaibeh, Texas Tech University
Measuring the Phylogenetic Nonrandomness of Biological
Data Sets
W.H.E. Day, Memorial University of Newfoundland
Fitting Trees to Biomedical Databases
Bill Shannon, Washington University at St. Louis
Performance of the Generalized Triples Algorithm for
Fitting Additive Trees
James E. Corter, Columbia University
Floor Discussion
2:45- 3:00 Coffee
3:00- 5:00 Invited Session 2: PHYLOGENETIC INFERENCE
Organizer: David Banks, NIST
Bayesian Phylogenetic Inference Via Markov Chain
Monte Carlo
Michael Newton, University of Wisconsin
Character Congruence, Taxonomic Congruence, and
Total Congruence: A Global Approach to Validate
Phylogenetic Trees.
Francois-Joseph Lapointe, Universite de Montreal
Discussant: W.H.E. Day, Memorial University of Newfoundland
Contributed Session 2: METHODS IN MEDICINE
Chair: Bill Shannon, Washington University at St. Louis
Risk Stratification in Advanced Renal Cell Carcinoma
M. Mazumdar, Memorial Sloan-Kettering Cancer Center
Discriminant Analysis with Singular Covariance Matrices
Richard E. Strauss, Texas Tech University
Modeling Uncertainty in Latent Class Membership
Kevin G. Lynch, University of Pittsburgh
Application of Cluster Analysis and Genetic Function
Approximation to the Discovery of Anticancer Drugs
Leming M. Shi, NIH
Floor Discussion.
5:15- 6:15 Invited Session 3: DNA FINGERPRINTING
Organizer: Joseph Gastwirth, George Washington University
Issues That Have Arisen Regarding the Application of
Population Statistics in Forensic DNA Typing
Bruce Budowle, FBI Forensics Laboratory
Discussant: Stefan Leigh, NIST
6:30- 9:00 Cocktails, Banquet
Banquet Address: Twinkle, Twinkle Little Star
Joseph Kruskal, Bell Laboratories, Lucent Technologies
June 14: Conference Sessions
9:00-10:45 Invited Session 4: NEW PROBLEMS IN BIOMEDICINE
Organizers: John Nolan, American University, and
Timo Koski, Royal Institute of Technology
Classification and Modeling in Functional Magnetic
Resonance Imaging
Chris Genovese, Carnegie Mellon University
Bayesian Predictive Identification of Bacteria
Timo Koski, Royal Institute of Technology
Classification Problems in Genetics
John P. Nolan, American University
Floor Discussion
10:45-11:00 Coffee
11:00-12:00 Invited Session 4: STRANGE BEDFELLOWS
Chair: Phipps Arabie, Rutgers University
Report on the DIMACS Symposium
F. R. McMorris, University of Louisville
A Bayesian Method for the Location of Regulatory
Regions in DNA
Evelyn Crowley, Carnegie Mellon University
Contributed Session 3: NEW IDEAS IN CLASSICAL DOMAINS
Chair: Stephen Hirtle, University of Pittsburgh
Financial Applications of Self-Organizing Maps
Guido Deboeck, World Bank
Challenges for the Future of Document Clustering
David S. Dubin, University of Illinois at Urbana-Champaign
Floor Discussion
12:00- 1:30 Lunch
1:30- 2:45 Invited Session 5: ENVIRONMENTAL APPLICATIONS
Organizer: Lara Wolfson, University of Waterloo
Detecting Outliers from a Multivariate Mixture Distribution
Wayne Woodward, Southern Methodist University
A Subjective Approach to Environmental Sampling
Lara Wolfson, University of Waterloo
Discussant: Peter Bryant, University of Colorado at Denver
Contributed Session 4: MATHEMATICAL STRATEGIES
Chair: Douglas Carroll, Rutgers University
The Crumpled Napkin Problem
Robert Olszewski, Carnegie Mellon University
An Upper Bound of Medians in Upper Semimodular Lattices
Jinlu Li, Shawnee State University
Multivariate Calibration with Partial Least-Squares
Regression for Mixture Analysis of Membrane
Introduction Mass Spectrometry Data
Susan L. Rose-Pehrsson, Naval Research Laboratory
Floor Discussion
2:45- 3:00 Coffee
3:00- 5:00 Invited Session 6: ISSUES IN SYSTEM SECURITY
Chair: Mike Windham, University of South Alabama
Fraud and Intrusion Detection in Financial Information
Systems: An Application of JAM (Java Agents for
Meta-Learning)
Sal Stolfo, Columbia University
Anomaly Detection in Process Control and System
Security
Roy Maxion, Carnegie Mellon University
Discussant: David Banks, NIST
Contributed Session 5: CLUSTERING METHODS
Chair: Glenn Milligan, Ohio State University
The MCLUST Project for Model-Based Classification
Chris Fraley, University of Washington and
MathSoft, Inc.
An Interior Point Algorithm for Minimum Sum-of-Squares
Clustering
Pierre Hansen, Universite de Montreal
The Validation of a Statistical Cluster Validation
Methodology with Empirical Data
Mary Ann Statman, The Human Resources Research Organization
Exact Algorithms for Additive Clustering
Christophe Meyer, Universite de Montreal
Combinatorial Statistical Tests for Clusters
Bernard Harris, University of Wisconsin
Floor Discussion
June 15: Conference Sessions
9:00-10:30 Invited Session 7: CLUSTERING PROBLEMS IN MARKETING
Organizer: Paul Green, University of Pennsylvania
Empirical Findings Obtained from Evaluating k-Modes
and Overlapping k-Centroids Clustering
Anil Chaturvedi, Bell Labs
Cluster-Based Market Segmentation: Some Further
Comparisons of Alternative Approaches
Catherine M. Schaffer, University of Pennsylvania
Discussant: Frank Carmone, Wright State University
10:30-10:45 Coffee
10:45-12:00 Invited Session 8: APPLICATIONS IN IMAGING
Organizer: David Crosby, American University
Image Enhancement in Blind Deconvolution and Parameter
Selection Using Informational Complexity
Hamparsum Bozdogan, University of Tennessee
Classification Techniques in Satellite Meteorology
David Crosby, American University
Discussant: TBA
Contributed Session 6: STRATEGIES FOR NEURAL NETS
Chair: Olga Cordero-Brana, American University
An Improved Probabilistic Neural Network for Chemical
Sensor Array Pattern Recognition
Ronald E. Shaffer, Naval Research Laboratory
Strategies for the Analysis of Superlarge Datasets
David Banks, NIST
Determining The Most Probable Number of Groups and
Outlier Detection
A. Ramaswami
Model Selection in Gaussian Mixtures Using Mean Square Error
Shibo Zheng, Anglia Polytechnic University
Floor Discussion