************************************************************* DIMACS Workshop on Data Depth: Robust Multivariate Analysis, Computational Geometry and Applications May 14 - 16, 2003 DIMACS Center, Rutgers University, Piscataway, NJ Organizers: Regina Liu Rutgers University, rliu at stat.rutgers.edu Robert Serfling University of Texas at Dallas serfling at utdallas.edu Diane Souvaine Tufts University dls at eecs.tufts.edu Yehuda Vardi Rutgers University vardi at stat.rutgers.edu Presented under the auspices of the DIMACS Special Focus on Data Analysis and Mining and the DIMACS Special Focus on Computational Geometry and Applications. ****************************************************** Multivariate statistical methodology plays a role of ever increasing importance in real life applications, which typically entail a host of interrelated variables. Simple extensions of univariate statistics to the multivariate setting do not properly capture the higher-dimensional features of multivariate data, nor do they yield geometric solutions because of the absence of a natural order for multidimensional Euclidean space. A more promising approach is the one based on "data depth", which can provide a center-outward ordering of points in Euclidean space of any dimension. Extensive developments in recent years have generated many attractive depth-based tools for multivariate data analysis, with a wide range of applications. The diversity in approaches, emphases, and concepts, however, makes it necessary to seek unified views and perspectives that would guide the further development of the depth-based approach. The concept of data depth provides new perspectives to probabilistic as well as computational geometries. In particular, the development of implementable computing algorithms for depth-based statistics has brought about many new challenges in computational geometry. This workshop would create a unique environment for multidisciplinary collaboration among computer scientists, theoretical and applied statisticians, and data analysts. It would bring together active researchers in these fields to discuss significant open issues, establish perspective on applications, and set directions for further research. ****************************************************** STUDENT SCHOLARSHIPS: The workshop offers some student scholarships to reimburse expenses for travel, registration, and lodging at the workshop, up to $400. Applicants for student scholarships should contact Regina Liu (rliu at stat.rutgers.edu). ****************************************************** Wednesday, May 14, 2003 8:00 - 9:00 Breakfast and registration 9:00 - 9:10 Opening remarks Fred Roberts, Director of DIMACS 9:10 - 9:20 Yehuda Vardi Department of Statistics, Rutgers University 9:20 - 9:40 Marianthi Markatou Program Director, NSF Statistics Program Chair: Arthur Cohen 9:40 - 10:40 An Overview on Depth Functions in Nonparametric Multivariate Analysis Robert Serfling, University of Texas at Dallas 10:40 - 11:00 Short Break Chair: Diane Souvaine 11:00 - 12:00 Depth and Arrangements David Eppstein, University of California, Irvine 12:00 - 1:30 Lunch Chair: Xuming He 1:30 - 2:10 Data Analysis by Zonoid Depth Karl Mosler, Universitat zu Koln, Germany 2:10 - 2:50 Data Depth in Multivariate Analysis: Dependence, Discrimination and Clustering Mario Romanazzi, "Ca' Foscari University of Venice, Italy 2:50 - 3:10 Fast Algorithms for Frames and Point Depth in Polyhedral Hulls Jose Dula, University of Mississippi and the U.S. Census Bureau 3:10 - 3:40 Break Chair: David Madigan 3:40 - 4:20 Maximal Depth Estimators of Regression Quantiles for Censored Data Steve Portnoy, University of Illinois 4:20 - 5:00 Data Depth and Mixture Models Ryan Elmore, Thomas Hettmansperger*, Fengjuan Xuan Penn State University 5:00 - 5:40 Comparing Multivariate Scale Using Data Depth: Testing for Expansion or Contraction Kesar Singh, Rutgers University 5:40 - 6:00 A Test for Equal Variances Using Data Depth Karen McGaughey, Kansas State University 6:00 - 7:00 RECEPTION (Wine and Cheese) Thursday, May 15, 2003 8:30 - 9:00 Breakfast and registration Chair: Marshall Bern 9:00 - 9:40 Real-time Computation of Data Depth Using the Graphics Pipeline Suresh Venkatasubramanian, AT&T Labs- Research 9:40 - 10:20 Computing the Center of Area of a Convex Polygon Pat Morin, Carleton University, Canada 10:20 - 10:40 Distance Problems on points and lines Ovidiu Daescu, Ningfang Mi, University of Texas at Dallas 10:40 - 11:00 Short Break Chair: Cun-Hui Zhang 11:00 - 11:40 Computation of Projection Depth and Related Estimators Yijun Zuo, Michigan State University 11:40 - 12:20 On some Probabilistic Algoritms for Computing Tukey's Half Space Depth Biman Chakraborty*, Probal Chaudhuri 12:20 - 2:00 Lunch Chair: Evarist Gine 2:00 - 2:40 On data based distances between data vectors and between hyperplanes Hannu Oja, University of Jyvaskyla, Finland 2:40 - 3:20 On the Computation and Robustness of Some Data Depth Notions Greg Aloupis, McGill University 3:20 - 4:00 Yehuda Vardi Deep Talk 4:00 - 4:30 Break Chair: David Tyler 4:30 - 4:50 A definition of depth for functional observations Sara López-Pintado, Juan Romo* Universidad Carlos III de Madrid, Spain 4:50 - 5:10 A Depth_Based Kurtosis Functional Jin Wang, University of Texas at Dallas 5:10 - 5:50 Tukey depth-based trimmed means Jean-Claude Masse, Universite Laval, Quebec, Canada 6:20 DINNER, DIMACS Lounge Friday, May 16, 2003 8:30 - 9:00 Breakfast and registration Chair: William Steiger 9:00 - 9:40 Clustering and Cluster validation via the Relative Data Depth Rebecka Jornsten, Rutgers University 9:40 - 10:20 On aspects of regression depth and methods based on convex risk minimization for data mining Andreas Christmann, University of Dortmund, Germany 10:20 - 10:40 Short Break Chair: Suneeta Ramaswami 10:40 - 11:20 Nonparametric clustering of high-dimensional data Bogdan Georgescu, Rutgers University Ilan Shimshoni, Technion-Israel Institute of Technology, Israel Peter Meer*, Rutgers University 11:20 - 11:40 Functional Samples and Bootstrap for Predicting SO_2 Levels Belen Fernandez-de-Castro*, S.Guillas, W. Gonzalez Manteiga Universidade de Santiago de Compostela, Spain 11:40 - 12:00 Exact, Adaptive, Parallel Algorithms for Data Depth Problems Vera Rosta, McGill University, Canada 12:00 - 1:20 Lunch Chair: Peter Rousseeuw 1:20 - 2:00 Analyzing The Number of Samples Required for an Approximate Monte-Carlo LMS Line Estimator David M. Mount, University of Maryland 2:00 - 2:40 Optimizing Depth Functions Stefan Langerman, Universite Libre de Bruxelles , 2:40 - 3:00 Short Break Chair: Peter Rousseeuw 3:00 - 3:40 A note on Simplicial Depth & Software Demonstration Michael Burr*, Eynat Rafalin*, Diane Souvaine, Tufts University 3:40 - 5:00 Open Discussion 5:00 Workshop adjourn ****************************************************** Registration Fees: (Pre-registration deadline: May 7, 2003) Regular rate Preregister before deadline $100/day After preregistration deadline $120/day Reduced Rate* Preregister before deadline $50/day After preregistration deadline $60/day Postdocs Preregister before deadline $10/day After preregistration deadline $15/day DIMACS Postdocs $0 Non-Local Graduate & Undergraduate students Preregister before deadline $5/day After preregistration deadline $10/day Local Graduate & Undergraduate students $0 (Rutgers & Princeton) DIMACS partner institution employees** $0 DIMACS long-term visitors*** $0 Registration fee to be collected on site, cash, check, VISA/Mastercard accepted. Our funding agencies require that we charge a registration fee for the workshop. Registration fees cover participation in the workshop, all workshop materials, breakfast, lunch, breaks, and any scheduled social events (if applicable). * College/University faculty and employees of non-profit organizations will automatically receive the reduced rate. Other participants may apply for a reduction of fees. They should email their request for the reduced fee to the Workshop Coordinator at workshop at dimacs.rutgers.edu. Include your name, the Institution you work for, your job title and a brief explanation of your situation. All requests for reduced rates must be received before the preregistration deadline. You will promptly be notified as to the decision about it. ** Fees for employees of DIMACS partner institutions are waived. DIMACS partner institutions are: Rutgers University, Princeton University, AT&T Labs - Research, Bell Labs, NEC Laboratories America and Telcordia Technologies. Fees for employees of DIMACS affiliate members Avaya Labs, IBM Research and Microsoft Research are also waived. ***DIMACS long-term visitors who are in residence at DIMACS for two or more weeks inclusive of dates of workshop. *************************************************************** Information on participation, registration, accommodations, and travel can be found at: http://dimacs.rutgers.edu/Workshops/Depth/ **PLEASE BE SURE TO PRE-REGISTER EARLY** ***************************************************************