|
Friends,
Please forward this invitation to your colleagues, and RSVP to
me (
This e-mail address is being protected from spam bots, you need JavaScript enabled to view it
) if you plan to attend.
The Statistics & Reliability Discussion Group cordially invites you to
attend an outstanding FREE presentation:
Speaker: Mr. Ed Russell
Topic: Selecting Winners: Robust Rating and Ranking in Busines Editorial Comment: If you are involved with project selection and
management you need to attend this talk. If you have to make
rational choices among several options be they personal, financial,
or whatever, you need to attend this talk. Ed has made many
outstanding presentations to our group and this is no exception.
John Flaig Abstract:
Knowing where you stand among your competitors is obviously
critical in sports. It's generally obvious that to be a winner,
you don't need to be the best at everything and to be a loser,
you don't need to be the worst at everything. In many sports,
like running or speedskating, knowing where you stand is fairly
simple. You just track your time and that of your competitors.
Things get more complicated in some sports, like: gymnastics,
snowboarding, and (horrors) figure skating. You might think
you just have to track your scores and those of your opponents –
but different panels may come to different conclusions. Worse,
we often suspect the same panel may give much different
conclusions on different days, raising that ugliest of all questions:
"Does the score actually mean anything?"
Life seems to become even more complex in selecting projects
to fund in business. There aren't any panels Judges who at least
had some training in how to declare winners. You sometimes
suspect that half the people involved in the selections in business
don't fully understand the choices. You are often pretty sure
that a few people are there to make sure their project wins approval
and that they just might kill yours to ensure theirs wins funding.
Perhaps one of the company Execs is in the room too. Does that
automatically turn the selection process into a divination session
where the goal is to figure out what (s)he wants?
Is there any methodology which can provide help in the seemingly
hopeless task? Will it work with multiple metrics? Is it possible to
use untrained "Judges" and still make a decent choice? Is there a
way to tell at least what is an OK choice from a really bad one?
We will approach this topic starting with gaining an understanding
of some of the basic issues in determining winners and losers in
the context of selecting projects to fund. We will compare a few
systems commonly used in business meetings to select projects.
We will mention one system, the Borda system, which is reasonably
well known in academics but not often used in business. And we
will introduce another system which is not very well known either
in business or academics, I will call it the Foldvary system, which
has some unique properties. Finally, we will make recommendations
on which system to use, when.
If we have time, we will make some comments on how these systems
can be used to help understand what is driving business performance
in the consumer marketplace.
Biography:
Ed Russell manages Statistical Modeling at National Semiconductor
where he primarily supports the analysis and comparison of SPICE
Models, Design Rules, and electrical test data.
Prior to working for National, Ed served as a Black Belt Statistician
at Sun Microsystems developing and teaching statistical design
methodology for 65 and 45nm CMOS circuit design. In the semiconductor
industry, Ed has also served as: Reliability Director and Manager of
Statistical Information systems for Cypress Semiconductors; Senior
Manager for Business Process Programs, Quality Manager, and
Statistician for AMD; and Manager of Statistical Programs for Sematech.
Earlier in his career, Ed has worked in various roles as: a geophysicist,
a statistician, and as a systems analyst / programmer.
Ed received an MS in Mathematical Statistics from Purdue University
with a focus on Decision Theory and Test of Hypothesis. In addition,
he has a formal background in: Biostatistics at the University of
Washington, Multivariate Analysis at Ohio State University, and
Geophysics at the University of Houston. Ed holds 2 patents and
has several additional patent applications in process
Please RSVP
This e-mail address is being protected from spam bots, you need JavaScript enabled to view it
at least two days prior to the
meeting if possible. |
