Fall 2001

• Matrix inequalities with applications to statistics.
  -- September 19, 2001 -- Wang Song Gui, Beijing Polytechnic U.

• Some Bayesian directions for likelihood based diagnostics.
  -- September 26, 2001 -- Michael Brimacombe, UMDNJ

• Sinc quadrature rules for univariate tail probabilities.
  -- October 5, 2001 -- FRIDAY -- Robert Strawderman, Cornell U.

• Selected topics in mixture models.
  -- October 10, 2001 -- Bruce Lindsay, Penn State U.

• Understanding high dimensional classifiers
  -- October 12, 2001 -- FRIDAY A.M. -- Ofer Melnik, Rutgers and AT&T Labs

• Experiencing information systems through sound.
  -- October 17, 2001 -- Mark Hansen, Bell Labs, Murray Hill

• Empirical and generalized Bayes ridge regression estimators with minimaxity and stability.
  -- October 24, 2001 -- Tatsuya Kubokawa, U. Tokyo

• Statistics and the evolution of the Internet.
  -- October 31, 2001 -- Andrew Odlyzko, Digital Technology Center, U. Minnesota

• Statistics and wine: An opportunity for a highly significant positive interaction.
  -- November 7, 2001 -- Simon Sheather, U. New South Wales and U. Sydney

• Strong admissibility, recurrent Markov chains and estimation in high dimensional normal mean estimation problems.
  -- November 14, 2001 -- Joe Eaton, U. Minnesota

• Confidence-Distributions: Concept, theory and applications.
  -- November 28, 2001 -- Kesar Singh, Rutgers University

• Large randomization designs in clinical trials and classical probability distributions.
  -- December 5, 2001 -- Andrew Rukhin, U. Maryland

Spring 2002

• Compatible prior distributions for DAG-models.
  -- January 23, 2002 -- Guido Consonni, University of Pavia (Italy)

• Multiple comparisons for making decisions.
  -- January 30, 2002 -- Jason C. Hsu, The Ohio State University

• Conjugate Bayesian inference for Gaussian graphical models.
  -- February 6, 2002 -- Alberto Roverato, University of Modena and Reggio Emilia

• Network tomography using the EM algorithm.
  -- February 13, 2002 -- Robert Nowak, Rice University

• Optimal predictive model selection.
  -- February 20, 2002 -- James Berger, Duke University

• The algebra of uniformly minimum variance unbiased estimators (UMVUEs) from categorical data.
  -- February 27, 2002 -- Abram Kagan, University of Maryland-College Park

• Joint Statistics-Computer Science Talk: The boosting approach to machine learning.
  -- March 4, 2002 -- Robert Schapire, AT and T Labs

• The square of an isotropically random unitary matrix is not isotropically random.
  -- March 6, 2002 -- Thomas L. Marzetta, Mathematical Sciences Research Center, Bell Labs, Lucent Technologies

• Maxima of Gaussian random fields and moderate deviation approximations to boundary crossing probabilities for sums of random variables with multidimensional indices.
  -- March 13, 2002 -- Hock Peng Chang, National University of Singapore

• Monotone functions: nonstandard limit theory for likelihood ratio statistic.
  -- April 3, 2002 -- Moulinath Banerjee, University of Michigan

• A type of restricted maximum likelihood estimator of variance components in generalized linear mixed models.
  -- April 10, 2002 -- Jason Liao, University of Medicine and Dentistry of New Jersey

• Robust inference for proportional hazards frailty regression models.
  -- April 17, 2002 -- Michael Kosorok, University of Wisconsin-Madison

• "I think they're pretty much the same; what do you think?"
  Statistical comparison of Nuclear Magnetic Resonance Spectra.
  -- April 24, 2002 -- Richard Spencer, NIH