[Cowystats] COWY Stats ASA Notes and News
pocernic at rap.ucar.edu
Fri Mar 30 12:52:21 MDT 2007
## April 11 - 13th Dr. Jianqing Fan - CSU Visit details
## April 20 Chapter Spring Meeting - NCAR, Boulder
*** Invited speaker abstracts
*** Other information
## April 27 Rick Katz - School of Mines
Dr. Jianqing Fan "Fredrick L. Moore" Professor of Finance, Princton
Univeristy and President elect of Institute of Mathematical Standards,
will be visiting CSU for several days and giving a series of talks.
The titles are below. Further information can be found at CSU's
High-dimensional statistcal learning and inference
Wednesday, April 11, 2007
A205 Clark 3:05 p.m.-4:00 p.m.
Statistical Analysis of DNA Microarray Data
Thursday, April 12, 2007
A202 Clark 3:05 p.m.-4:00 p.m.
Option pricing with aggregation of physical models and statistical learning
Friday, April 13, 2007
Glover 130 3:00 p.m.-4:00 p.m.
Details can be found at
###Colorado/Wyoming ASA Chapter Spring Meeting
Friday April 20th 9am to 4:30.
National Center for Atmospheric Research Mesa Lab.
Possibly the most beautiful places to have a meeting on the front range.
Plan now. Our spring meeting is 3 weeks from today. The event is free
and advanced registration is not required. Perfect weather
guaranteed. Below are some abstracts from the invited speakers. A
schedule of the day's events is still being created. At the spring
meeting we like to offer students a chance to present some of their
research. This is typically arranged through the schools, but if
there are students interested in giving a short presentation, they are
encouraged to contact me at pocernic at ucar.edu.
This year there are three Chapter offices open for election:
secretary, newsletter editor and president-elect. In recent years,
the officers have been operating more like an executive board. We
meeting a couple times a year but most business using emails. The
newsletter has been replaced by these emails notes and we have been
using meeting agendas to summarize meetings. A person serves as
president-elect for a year before becoming president for a year.
The officers elected will have some additional duties and
opportunities with the JSM being here in Denver next summer.
If you or someone you know might be interested in being an officer,
please let us know. The obligations aren't overwhelming but they do
require a steady commitment.
**** Tour of NCAR's visualization lab*
Back by popular demand, a demonstration of NCAR's VisLab will occur
before the meeting. The VisLab is a state-of-the-art scientific
visualization environment, providing an immersive environment for
visualizing complex datasets in stereo-3D and collaborating across
sites via AccessGrid video teleconferencing. From a statistical
perspective, the VisLab allows data to be illustrated with both motion
and depth. A 30 minute demonstration will be held before the meeting.
More information on the VisLab can be found at
Selected Speaker Abstracts
Multiple Comparisons in Clinical Trials for Regulatory Purposes: A Brief
Overview with Discussion
Brian L. Wiens
Gilead Colorado, Inc.
We consider the problem of multiple hypothesis tests in clinical trials
aimed at supporting regulatory approval of a new medical therapeutic.
Sponsors (generally pharmaceutical companies) must plan a multiple
comparison procedure that allows for precise differentiation of an
investigational product compared to placebo. This often begins with
specification of one or more primary endpoints, one or more secondary
endpoints, etc. This can be complicated by the use of multiple treatment
groups such as multiple doses or multiple regimens of the investigational
product. The current trend seems to be toward using more complicated study
designs due to factors such as increased sophistication of clinical trial
specialists and a desire for faster, more efficient drug development. We
discuss the regulatory hurdles and some ways that statisticians are
approaching them, including some state-of-the-art procedures for handling
multiple comparisons such as fallback, gatekeeping and tree-structured
Covariance Tapering for Likelihood Based Estimation in Large Spatial Datasets
Cari Kaufman, Statistical and Applied Mathematical Sciences Institute
(SAMSI) and National Center for Atmospheric Research (NCAR)
Likelihood-based methods such as maximum likelihood, REML, and Bayesian
methods are attractive approaches to estimating covariance parameters in
spatial models based on Gaussian processes. Finding such estimates can be
computationally infeasible for large datasets, however, requiring O(n^3)
calculations for each evaluation of the likelihood based on n observations.
I will discuss the method of covariance tapering to approximate the
likelihood in this setting. In this approach, covariance matrices are
``tapered,'' or multiplied element-wise by a compactly supported correlation
matrix. This produces matrices which can be be manipulated using more
efficient sparse matrix algorithms. I will present two approximations to
the Gaussian likelihood using tapering and discuss the asymptotic behavior
of estimators maximizing these approximations. I will also present an
example of using the approximations in a Bayesian model for the
climatological (long-run mean) temperature difference between two sets of
output from a computer model of global climate, run under two different land
Fiducial Inference of R. A. Fisher -- History, Applications, and
Speaker: Hari Iyer, Department of Statistics, Colorado State University
R. A. Fisher introduced the FIDUCIAL ARGUMENT which he used to derive
confidence intervals for the difference between two normal means when
the variances are unequal -- the so called Behrens-Fisher problem. He
also used this approach to derive interval estimates in several other
situations. However his contemporaries found a number of shortcomings
with the fiducial method. It is safe to say that the fiducial method
eventually fell into disfavor among statisticians. Recently the fiducial
method appears to be enjoying a revivial of sorts. In this talk I will
give a brief overview of the history of fiducial inference, give
examples and applications of the fiducial argument, and discuss recent
generalizations that makes the fiducial approach a powerful tool for
deriving inference procedures. Asymptotic properties and small sample
simulations confirm that fiducial procedures have excellent operating
characteristics in general. [Based on joint work with my
colleagues Jan Hannig, Jack Wang, and several former and current PhD
students from the department of statistics at Colorado state university].
Rick Katz - NCAR
Colorado School of Mines
Friday Afternoon April 27
Details to follow
An Archive of these Chapter Announcements can be found at
Have a good weekend.
National Center for Atmospheric Research
Research Applications Laboratory
More information about the Cowystats