2002-2003
September 6, 2002
Title: "Prophet-type inequalities for optimal
stopping with a random number of observations"
Pieter Allaart, University of North Texas
September 27, 2002
Title: "How I spent my summer watching TV"
Michael Monticino, University of North Texas
Abstract: So much of TV is bad. Why? Don't
television networks do market research? The answer is yes
they do, but they don't necessarily listen to the numbers.
This talk will discuss market research analysis conducted
for a major TV network.
October 4, 2002
Title: "Are you a good risk? Matching credit
risk to lending rates"
Michael Monticino, University of North Texas
Abstract: So you are trying to get a loan?
Why did the bank offer you the rate it did? The rate you
are offered is a direct result of how risky you appear to
the bank. Your risk used to be determined by a human "expert"
working at the bank. However, lending institutions are becoming
much more sophisticated in assessing and matching risk to
lending rates. Automated underwriting based on statistical
models is becoming more common. This talk will introduce
basic notions of credit scoring and credit scores are used
to determine loan approval and rates.
November 8, 2002
Title: "Understanding the Stochastic Nature
of p-values to Interpret
Hypothesis Tests"
Dr. Robert Pavur, University of North Texas, BCIS Department,
division of Management Science.
Abstract: The p-value is one of the most
commonly used statistical measures that is reported in research
papers. The talk will examine the use of p-values to assess
the performance of statistical tests. The presentation will
also comment on the use of a mixture of density functions
to model the distribution of p-values.
February 14, 2003
Title: "Properties of Gaussian Random Field
Models"
John Quintanilla, University of North Texas
Abstract: This is the introductory talk
of a three-seminar series concerning recent modeling of
aerogels using a Gaussian random field model. In this first
talk, we define Gaussian random fields and study the microstructure
of their excursion sets; this model has been employed in
the literature for many different materials. We then use
concepts from renewal theory to compute the autocorrelation
and chord-length density functions for this model.
February 21, 2003
Title: "Scattering and Modeling Methods to
Characterize Porous Materials"
Rick Reidy
Abstract: Scattering methods are very useful
tools to describe the structure and in some cases the composition
of materials. Because scattering data is based on distributions,
modeling is necessary to provide a description of the scattered
body. This discussion will compare scattering versus imaging
techniques such as electron microscopy, provide background
into the importance of materials characterization, describe
several modeling methods, and highlight the importance of
recent work done by Dr. Quintanilla et. al.
2001-2002
October 13
Title: "A Survey of Panel Tests for Stationarity
in the Presence of Structural Change with an Application
to Unemployment"
Margie A. Tieslau, University of North Texas, Department
of Economics
October 26
Title: "Evaluation of a Proposed Goodness-of-Fit
Test for a Poisson
Process"
Dr. Robert Pavur, UNT Department of BCIS
November 2
Title: "How to make a retirement fund last
forever"
John Quintanilla, University of North Texas
November 16
Title: "The first digit phenomenon"
Pieter Allaart, University of North Texas
November 30
Michael Monticino, University of North Texas
January 25
Title: "How to stop well on the average"
Pieter Allaart, University of North Texas
February 5, 2002
Michael Monticino, University of North Texas
February 8, 2002
Title: "Some remarks about Stochastic Cellular
Automata"
Thomas Böhme, University of North Texas, Computer Science
February 15, 2002
Title: "An elementary proof of Pearson's chi-square
test"
John Quintanilla, University of North Texas
April 26, 2002
Title: "Several Issues in Evolutionary Algorithms
and Genetic Algorithms"
Adam Kinney, University of North Texas
Abstract: Evolutionary Algorithms are
a broad class of probabilistic optimization techniques inspired
by biological adaptive systems, and Genetic Algorithms are
the most popular type of Evolutionary Algorithm. I will
outline the essential elements of an Evolutionary Algorithm,
and I will define the specific characteristics of a Genetic
Algorithm. Also, I will discuss some data representation
issues and alternative genetic operators in Genetic Algorithms,
as well as my plans for future investigations in Evolutionary
Algorithms.
