[Cowystats] CO/WY ASA Note Nov 14.

[Matt Pocernich]

### Jery Stedinger talks - Colorado State University, November 15, 16
###  Job Opportunity - Stratavia

### Jery Stedinger - PRIMES Lecture

Jery Stedinger from Cornell University will be giving a series of
lectures this week at CSU in Fort Collins as part of the PRIMES
distinguished lecture series.  Stedinger is equally prominent in the
field of hydrology as well as statistics.  The two talks pertain to
the use of extreme value theory as applied to estimating the return
level of floods.

Distinguished Lecture: Frequency Analysis for Extreme Hydrologic
Events:  Annual Floods and Precipitation
Thursday, November 15, LSC Senate Chambers, 3-4 p.m.

The three-parameter Generalized Extreme Value (GEV) distribution has
found wide application for describing
annual floods, rainfall, wind speeds, wave heights, snow depths and
other maxima. Hosking and Wallis (1997)
popularized use of L-moments and the GEV distribution as the basis of
an index-flood regional L-moment
flood-frequency estimation procedure. L-moments are linear
combinations of the ordered observations, and provide
nearly unbiased estimators of the L-coefficients of variation and
skewness. L-moment can also be used of
distribution selection. Recognition of the value of regionalization
procedures has been one of the great advances in
flood frequency analysis in the last two decades.

Regional Analysis of Hydrologic Data with Bayesian Generalized Least Squares

Friday November 16, 3:00 PM at Room 224-226 CSU Lory Student Center

Hydrologists often need to estimate hydrologic quantities, such as the
annual mean flow, flood
quantiles and low-flow statistics for water resources planning and
floodplain management, at
ungauged sites using regional information. Reis et al. (2005)
introduced a quasi-analytic
Bayesian analysis for Generalized Least Squares (GLS) regression which
has several advantages
over widely used estimators developed in earlier work (Tasker and
Stedinger, 1985, 1989). The
Bayesian approach provides both a measure of precision of the model
error variance that earlier
analysis procedures lacked, and a more reasonable description of the
model error variance in
cases where Maximum Likelihood Estimators (MLE) and Method of Moments
(MOM) model
error variance estimators are zero.

More information about both lectures can be found at

http://www.primes.colostate.edu/Quick%20Links/Calendar/PRIMES%20Calendar%20fall%202007.htm

### Job Opportunity  Strativaia

For further information contact Rick Osborne at  rick.osborne at gmail.com
Primary Responsibilities:

 - Work alone or with a small team of mathematicians to understand the
 feasibility and most optimal solution alternatives for proposed
analytics features.
 - Using mathematical toolkits like Matlab, R or others, prototype
solution alternatives
 - Document formulas, algorithms and other mathematical artifacts that
 will be handed off to software engineering for eventual incorporation
int our Data Palette product.
 - Work with software engineer, quality engineering, documentation and
 other Engineering departments to ensure the successful incorporation of
the analytics feature into the product.
 - Provide experienced leadership and mentoring to other, less
 experienced Analytics team members.
 - Collaborate with Product Marketing to evaluate the market place and
to contribute ideas for future analytics product direction.

Minimum requirements:

 - Minimum BA/BS (Masters or PhD preferred) in Statistics or related
 field having significant statistics content
 - A minimum of 3 years industry experience providing mathematics
 solutions for production products
 - Proven experience in the following areas:
 - Data Mining
 - Time series analysis techniques
 - Smoothing / denoising
 - Autoregressive Processes (AR), Moving Average Processes (MA), ARMA, ARIMA
 - Nonstationary and Seasonal Time Series Models
 - Autocovariance / autocorrelation / partial autocorrelation

Other qualities:

 - Proven oral and written communication skills
 - Able to work productively alone or in group environments
 - Experience with iterative or other software lifecycle methodologies