[Statlist] next talk on statistics

Christina Kuenzli kuenz|| @end|ng |rom @t@t@m@th@ethz@ch
Mon Jan 17 09:31:54 CET 2005


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

**************************************************************************
 
           We are pleased to announce the following talk
**************************************************************************
Friday, January 28, 2005, 15.15, LEO C15

        Pitfalls in parameter estimation through numerical optimization

        Douglas Bates, Department of Statistics, 
                       University of Wisconsin, Madison}  

Frequently the parameter estimates in statistical models are defined
as those values that optimize a particular objective function.  For
example, the maximum likelihood estimates are the parameter values
that maximize the likelihood function (or, equivalently, maximize the
log-likelihood function).  Determining the estimates and assessing the
variability in such estimates for a particular model/data set
combination often requires numerical optimization of the objective.
Although the theory of numerical optimization usually deals with just
the iterative optimization step, we consider the process to have three
important stages: determination of starting values for the parameters,
iterative optimization, and assessing convergence.  All three of these
stages are important to successful parameter estimation.  In addition,
careful consideration of the form of the parameterization and the form
of any constraints on the parameters facilitates the process.  We
illustrate with examples from nonlinear regression and from multilevel
modeling.

*************************************************************************

(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 using stat.math.ethz.ch>
Seminar fuer Statistik      
Leonhardstr. 27,  LEO D11          phone: +41 1 632 3438         
ETH-Zentrum,                       fax  : +41 1 632 1228 
CH-8092 Zurich, Switzerland        http://stat.ethz.ch/~




More information about the Statlist mailing list