No Subject

[Domenico Marinucci <marinucc@scec.eco.uniroma1.it>]

***********************************************************************
SEMINARI DI PROBABILITA' E STATISTICA MATEMATICA

DIPARTIMENTO DI MATEMATICA - UNIVERSITA' LA SAPIENZA  di ROMA
***********************************************************************

giovedi' 4 luglio, ore 17 aula CONSIGLIO

Larry Wasserman  ( Dept of Statistics - Carnegie Mellon University)

False Discovery Rates: A Large Sample Approach

In certain applications, such as genetic microarrays,
source detection in astrophysics and analysis of fMRI brain images,
researchers need to test thousands or even millions of hypotheses.
Rather than controlling the familty wise error, it makes
more sense to control the false discovery rate (FDR) which
is simply the number of false rejections of the null hypothesis
divided by the number of rejections. Benjamini and Hochberg (1995)
introduced the idea of FDR and provided a significance threshold that
controls the expected value of FDR. Their method is based on
certain order statistic arguments. We take a different approach.
We treat the FDR as a stochastic process (as a function of the
significance threshold) and we apply large sample theory to the
process. This leads to new insights as well as new methods.
This is joint work with Chris Genovese.


giovedi' 4 luglio, ore 18 aula CONSIGLIO

Nikolai Leonenko (Cardiff University)

Minimum contrast estimation of random processes and fields based on
information of the second and third order

We present the results on consistency and asymptotic normality of a
class of minimum contrast estimates for stochastic processes and
fields with short- and long-range dependence based on the second and
third order cumulant spectra.The method does not require
discretization,which is necessary in existing approaches.These
results are obtained jointly with V.Anh and L.Sakhno.

----------------------------------------------------------------------------
-----------

 tutti gli interessati sono invitati a partecipare, 

G. Nappo