[Cowystats] ASA Chapter Information Feb part 1
Matt Pocernich
pocernic at rap.ucar.edu
Fri Feb 9 15:23:37 MST 2007
## Feb 12 Aparna V. Huzurbazar, CU Denver
## Feb 15 Donna Stroup, CU Denver
## Feb 28 Tom Hamill (NOAA) Improving Probabilistic Weather Forecasts
## Mar 9 Data mining and geospatial statistics at DU
## Mar 28 UCHSC Biometrics seminar (noon)
## Apr 5-7 Colorado High School Science Fair - statistics award
## Apr 20 CO/ WY Spring meeting
## Apr 27 Seminar Colorado School of Mines (tentative)
People complain about the weather, but without it, what would anyone
talk about.
Below are descriptions for a series of talks on a variety of topics.
While they are applied talks on specific topics, they deal with
larger statistical issues. The speakers are aware that their audience
have a range of statistical interests so don't be concerned that the
talks will deal with topic specific minutia. In short, we encourage
everyone to attend.
Also note that CU-Denver has two talks from visitors next week.
We are still seeking two additional volunteers for high school state
science fair. The primary time commitment for this in on April 6th in
Fort Collins.
Thanks,
Matt
##############################
Monday, February 12, 2007 1:30 PM - 3:00 PM
Dr. Aparna V. Huzurbazar, University of New Mexico
Flowgraph Models for Multistate Time-to-Event Data: Extensions and
Applications
CU-Denver Building, Room 480
1250 14th St. Denver, CO 80202
Flowgraph models are useful in a variety of reliability and survival
analysis problems. Previously, they have been used to model progression of
diseases such as cancer and AIDS, and for degenerative diseases such as
diabetic retinopathy, kidney failure, and dementia. They
are especially useful for analyzing time to event data and constructing
corresponding Bayes predictive distributions, survivor functions, and
hazard functions. Flowgraph models are general multistate stochastic models
that, when combined with transform inversion, allow a wide variety of
parametric time-to-event modelling.
################
Thursday, February 15, 2007 - 11:30 AM - 1:00 PM
Second Annual Joint CU-Denver/Colorado School of Mines Seminar
Dr. Donna Stroup, Executive Director, Data for Solutions, Inc
Statistical Considerations in Environmental Epidemiology: Assessing
Effects of Air Pollution on Human Health
CU-Denver Building, Room 656 1250 14th St.
We examine three important methodological considerations for analysts
developing and validating models to estimate the effect of air pollution,
especially particulate matter (PM), on the health of populations. Changes
in air pollution, particularly particulate matter (PM), can be associated
with changes in human health. In fact, this association may depend not only
on how air pollution is measured but also on important statistical and
methodological issues inherent in analytic modeling and analysis.
For directions and more complete abstracts, please visit
http://www-math.cudenver.edu/
http://www-math.cudenver.edu/contact/map2office.html
##################################
Wednesday February 28th - 4pm
Improving Probabilistic Weather Forecasts
Thomas M. Hamill
NOAA Earth System Research Lab, Physical Sciences Division
Boulder, CO
Metro State College Denver
Jointly sponsored by Metro Department of Meteorology, Metro Department of Statistics and CO/WY ASA.
Refreshments provided.
In the Science Building Room 103 Metro State College Denver
See http://www.mscd.edu/enroll/admissions/campus/map/ for a campus
and parking information. Note that the campus is very well served by
RTD light rail lines.
Abstract ***
Weather forecasts, we all know, inevitably have errors, sometimes small,
sometimes large. There are two main sources of errors in numerical
weather predictions. The first is "chaos," or the sensitive dependence
on the initial condition. Any small error in the initial description of
the weather will tend to grow larger and larger as the forecast lead
increases, though the error growth rate may vary from one day to the
next. The other source of error is imperfections in the numerical model
used to make the weather forecast.
While the current practice is to issue deterministic forecasts ("high of
60 today") a more intellectually honest way of communicating the
forecast is to do so probabilistically, indicating in some fashion the
uncertainty in the forecast.
In this talk we will demonstrate how a long time series of past
forecasts, which we call "reforecasts" and associated weather
observations can be used to generate skillful, reliable probabilistic
forecasts, forecasts that account for the uncertainty due to chaos,
forecasts that statistically adjust for many of the systematic errors
introduced by the forecast model. We'll show that a statistically
adjusted weather forecast is much more skillful than a forecast without
this statistical adjustment.
######################
Spatial Statistics and Data Mining
Friday March 9th 2 pm through 4:30
University of Denver
Details to follow soon.
######################
### Keith A. Baggerly from the University of Texas M. D. Anderson
Cancer Center Thurs March 29th: Preventive Medicine and Biometrics
Seminar 12-1pm CPH Auditorium 1J05, 9th St Campus, UCHSC sponsored
with the Bioinformatics Department at the University of Colorado
Health Science Center. (Abstract to follow)
######################
Volunteers needed to judge at the Colorado State
The Chapter will be offering a series of awards at the state science
fair for projects that exhibit good use and understanding of
statistics. The event is April 5 - 7 at Colorado State University in
Fort Collins. We need 2-3 people to help judge exhibits on Thursday
April 6 from approximately noon to 5 pm. If you are interested, send a
note to pocernic at ucar.edu.
######################
Spring Meeting Friday April 20th - NCAR Mesa Lab - beautiful
weather guaranteed. In addition to several invited speakers, we
strongly encourage students to present their research. Keep in mind
the meeting is a friendly environment in which to let the local
statistics community know what you have been working on the the last n
years. More details to follow.
#####################
Job opportunities - also see http://www-math.cudenver.edu/statistics/jobs.shtml
SigmaPro - Statistical software developer
SigmaPro is seeking a highly qualified software developer to help us
deliver successful products to the marketplace. The successful candidate
will be experienced in developing integrated applications in C# and Visual
Basic using Visual Studio as well as HTML. The position involves software
development, teaching and coaching clients on statistical methods, and
customer support. Desirable technical skills are in the areas of
multivariate statistical analysis, design of experiments, matrix
computations, simulation and optimization. Familiarity with Visual
Numerics IMSL libraries would be very advantageous.
More information can be found at
http://www.sigmapro.net/Resources/Employment/tabid/85/Default.aspx
Matt Pocernich
National Center for Atmospheric Research Research
Applications Laboratory (303) 497-8312
--
Matt Pocernich
National Center for Atmospheric Research
Research Applications Laboratory
(303) 497-8312
More information about the Cowystats
mailing list