Plenary Speaker:
Jeff Wu
TC.F. "Jeff" Wu
is a professor in ISyE and holds the Coca-Cola Chair in
Engineering Statistics. He was formerly the H. C. Carver Professor
of Statistics and Professor of Industrial and Operations
Engineering at the University of Michigan, Ann Arbor from 1993 -
July 2003, the GM/NSERC Chair in Quality and Productivity at the
University of Waterloo from 1988-1993, and before Waterloo, he
taught in the Statistics Department at the University of Wisconsin
from 1977-1988. He earned his BS in Mathematics from National
Taiwan University in 1971 and Ph.D. in Statistics from the
University of California, Berkeley (1973-1976). Dr. Wu joined
Georgia Tech in the summer of 2003.
Dr. Wu's accomplishments include receiving the prestigious COPSS
(Committee of Presidents of Statistical Societies) Presidents
Award in 1987 which is presented annually to the best researcher
under the age of 40. He has also been commissioned by five
statistical societies, elected a Member (Academician) of Academia
Sinica, and a Fellow of the American Society for Quality, of the
Institute of Mathematical Statistics and of the American
Statistical Association. Dr. Wu has won numerous awards, including
the 1990 Wilcoxon Prize for the best practical application paper
in Technometrics, the 1992 Brumbaugh Award for the single most
important paper to quality control among the publications
sponsored by the American Society for Quality Control, and the
1997 Jack Youden Prize for the best review paper in Technometrics.
He was the 1998 P. C. Mahalanobis Memorial Lecturer at the Indian
Statistical Institutes with widely cited research work and a
listing as an “ISI Highly Cited Researcher” in 2002 on
www.isihighlycited.com (ISI = Institute for Scientific
Information).
Dr. Wu's work is widely cited in professional journals as well as
in magazines, including a feature article about his work in
Canadian Business and a special issue of Newsweek on quality. He
has served as editor or associate editor for several prestigious
statistical journals like Annals of Statistics, Journal of
American Statistical Association, Technometrics, and Statistica
Sinica. Professor Wu has published more than 100 research articles
in peer review journals. He has supervised 29 Ph.D.'s, out of
which 14 are teaching in major research departments in
statistics/engineering/business in US/Canada and two are senior VP
in major US companies. He co-authors with Mike Hamada the
award-winning book "Experiments: Planning, Analysis, and Parameter
Design Optimization" (Wiley, 2000, 638 pages). |
A
Modern Theory of Factorial Designs
A modern theory of factorial designs is outlined that attempts to
encompass the major work in the last two decades. Traditionally,
regular orthogonal designs, include the 2^{n-k} and 3^{n-k} series,
have been used for factorial experiments. A central issue in
conducting factorial experiments is the optimal assignment of
factors and interactions to columns of the design. This is equivalent
to the choice of fractions according to some optimality criterion. The
minimum aberration (MA) criterion (and related criteria) has emerged
as the
major criterion for this problem. Major results and techniques,
especially the use of complementary sets, will be presented for
finding MA s^{n-k} designs, s being a prime power. The problem becomes
harder if the factors cannot be treated symmetrically (e.g., blocking
or split-plot structure, and robust parameter designs.) Extensions of
such results to nonregular orthogonal designs will be briefly
outlined. Many of the results in this talk are available from the
forthcoming book "A Modern Theory of Factorial Designs" by Rahul
Mukerjee and C. F. Jeff Wu (2000, Springer). |