[Statlist] Seminar ueber Statistik

[Christina Kuenzli]

                       ETH and University of Zurich 
      Proff. A.D. Barbour -- P. Buehlmann -- F. Hampel -- H.R. Kuensch 

**************************************************************************     
           We are pleased to announce the following talks
**************************************************************************

Friday, June 10, 2005, 15.15, LEO C15
     
      Semiparametric estimation in copula models

      Hideatsu Tsukahara, Seijo University, Tokyo 

**************************************************************************
Wednesday, June 22, 2005, 15.15, LEO C15

      Small Sample Statistical Modeling and Inferring Genomic
      Networks

      Juliane Schaefer, Dept. of Statistics, University of Munich

The advent of high-throughput experiments is promising with
respect to elucidating questions in molecular and genome
biology. However, due to the large number of investigated features as
compared to usually only a small number of samples, current genomic data
pose substantial challenges for statistical modeling and analysis. 

In my talk I will discuss the inference of large-scale gene relevance and
gene association networks from expression data. These require the
estimation of (inverse) covariance and correlation matrices in order to
describe gene interactions. This is a difficult task as traditional
estimators, albeit widely used in bioinformatics, turn out to be
inefficient when there are many variables but only few
observations. Instead, a regularized covariance matrix estimator needs to
be employed. In our approach to large-scale graphical Gaussian modeling we
propose to use variance-reduced and shrinkage-transformed estimators, along
with heuristic empirical Bayes model selection, in order to infer the
respective network topologies.
**************************************************************************

Friday, June 24, 2005, 15.15, LEO C15

      Algebraic Factor Analysis: Tetrads, Pentads and Beyond

      Bernd Sturmfels, University of California, Berkeley

Factor analysis is a classical technique for modeling correlated
observed variables as conditionally independent given few hidden
variables, the so-called factors. In this lecture, we consider Gaussian
random variables, and we discuss the polynomial constraints imposed
by factor analysis on the covariance matrix of the observed variables.
For the one-factor model these constraints are tetrads, for the
two-factor model they are the pentads and 3x3-determinants, etc...
We examine the geometry of this hierarchy, and we present some new
higher-degree constraints which were computed using Gr\"obner bases.
This is ongoing joint work with Mathias Drton and Seth Sullivant.
     
**************************************************************************      

(LEO (Leonhardstrasse 27, 8006 Zurich) is close to the main building,
across the hill-side station of the 'Polybahn') 
Overview maps of ETH : http://www.ethz.ch/search/orientation_en.asp
________________________________________________________
Christina Kuenzli            <kuenzli at stat.math.ethz.ch>
Seminar fuer Statistik      
Leonhardstr. 27,  LEO D11      phone: +41 (0)44 632 3438         
ETH-Zentrum,                   fax  : +41 (0)44 632 1228 
CH-8092 Zurich, Switzerland        http://stat.ethz.ch/~