News & Notes
Send in alumni updates or suggestions for future newsletters
I wish to thank Pam McGhee for the superb job she has done in creating the Newsletter. Pam whimsically suggested that I write this in Latin, but I have to admit that despite five years of it, almost nothing remains --- the only "foreign" languages I write with any confidence these days are computer languages.
My experiences as Chair over the last year have mostly been positive, and I can report that I have learned more (but never enough) than I ever expected to know about how a major research university operates. Generally, it is a pleasure to work on projects around campus and to see how our department contributes to the university as a whole. FSU was recently identified as the most efficiently run university in the nation, and I would like to think that this also reflects favorably on us.
There have been several notable successes for our faculty and students this year. Myles Hollander and Jayaram Sethuraman both won Professorial Excellence Awards. Myles also completed his term as editor of the Theory and Methods Section of the Journal of the American Statistical Association. Xufeng Niu was promoted to associate professor with tenure. Numerous research articles by our faculty and students have appeared in the best journals. Doctoral students Lihong Qi and Shaojun Zhang won University Fellowships. In August we welcomed a new faculty member, Anuj Srivastava. Anuj comes here after a year as a post-doc at Brown University, where he worked with Ulf Grenander. He is expected to significantly enhance the department's visibility in the area of computationally intensive methods. Ten new graduate students also joined the department in August.
Alumni visiting the department these days may be surprised to find that the entrance to the main office of the department has changed, and it now goes through what used to be a storage room! Amazingly, it works. The faculty offices in that area are now much quieter, and the whole operation of the main office is improved. The credit for this success goes to our Office Manager, Virginia Grice, who, with her usual enthusiasm, pushed through the plan from start to finish.
Finally, I want to mention how gratifying it is to see that no fewer than five of our graduate students are competing for the newly established Yongyuan and Anna Li Award for best student colloquium. The endowment for this award was made possible by a generous donation from Anna Li, in memory of her husband, who tragically died from cancer last April. Yongyuan was one of our most promising graduate students, and this award is a fitting memorial to his life.
One project on which Srivastava has done significant research is that of Automated Target Tracking and Recognition (ATR). Automated Target Tracking and Recognition involves the observation of objects (targets) by remote sensors such as video, radar, or infrared technology and the development of algorithms to analyze those objects for detection, tracking, and recognition. Srivastava’s research in the field of ATR involves using a Bayesian approach to generate stochastic inferences on complex scenes using Jump-Diffusion Processes. “Based on pattern theoretic representations of dynamic scenes involving rigid target motions, a posterior measure is defined on a countable union of connected subspaces, each subspace representing a hypothesized model. Then, using an Ergodic Markov process, composed of continuous paths and discrete jump transformations, this posterior is empirically simulated to evaluate the conditional means.”
The development of algorithms for automated recognition of objects has
been identified as an area of critical importance to the mission of the
Department of Defense. The technology involved in ATR also has applications
in the civilian sphere, especially in the medical field where image analysis,
in the form of X-rays, CT scans and MRI’s, plays a vital role. Concerning
his research in image analysis and ATR, Srivastava says, “The basic idea
is this: when we look at objects (say humans, animals, handwriting)
we recognize them immediately. The question is: ‘Can we design
a machine to do that efficiently?’ The answer so far has been ‘No.’
Designing computer algorithms for automated recognition of objects from
their images is still a distant goal. One of our goals is to understand
how the brain identifies objects, model that phenomenon mathematically,
computer code to do that. We believe that statistical inference has an important role to play in this task.”
Srivastava’s work in image analysis has earned him an invitation to consult with the Army Research Office’s Center for Imaging Sciences. He is currently teaching the course STA 5106, Computational Methods in Statistics I.
To satisfy the conditions of the phased retirement program, Meeter will
teach in the spring semesters until the year 2002. A few students
who have heard about his retirement have already changed their programs
of study so that they will be able to take certain classes when Meeter
will teach. Although he still comes into the office occasionally
when in town, Meeter seems to be thoroughly enjoying his newly found
free time. This June he visited Austria and Italy. Recently,
he returned from a
month-long trip to the mountains of North Carolina.
Fred Leysieffer’s career with Florida State has always been filled with service to the university. Since joining our department in 1964, he has served on over 38 university committees, been department chair for nine years, served as Acting Dean of Arts & Sciences for a year and as an associate dean of Arts & Sciences for three years. His service in teaching includes directing several doctoral students and instructing students in courses from the introductory level through the advanced graduate level.
With all of his responsibilities, Leysieffer’s commitment to the statistics profession has remained strong. At the 1996 annual meeting of the Florida Chapter of the American Statistical Association, the chapter recognized Dr. Leysieffer’s contributions to the profession by electing him as the recipient of the Chapter Service Recognition Award, an honor which the chapter gives to only one person every three years.
The department congratulates Fred Leysieffer on his appointment as the Assistant Vice President for Academic Affairs and on his continued success. We wish him well and hope that he will continue to enjoy his remarkable career.
One of Dr. Niu’s main research interests is that of environmental data
analysis -- a field which provides fertile ground for interdisciplinary
research projects. Since 1993 Niu has been working with Robert J. Livingston
in Biology on the multivariate time series analysis of shrimp, riverflow
and rainfall data. The joint project with Livingston studies the effects
of the reduction of freshwater flow on the biological productivity in the
Apalachicola Bay system, “which has a tremendous value to the state of
Florida.” In 1994, Niu began working with Ian McKeague and James B. Elsner (Meteorology) on “Empirically determined climate predictability using nonlinear time series models” and “Societal vulnerability to hurricanes” – projects which “pose many challenges to statisticians, especially the need for taking into account high-dimensional inputs/outputs, nonlinear covariate effects, spatial dependence and trend estimation.” Niu’s latest interdisciplinary project has been “Recruitment forecast models for economically important reef fishes in the eastern Gulf of Mexico” with Dr.’s Felicia Coleman and Christopher Koenig in Biology.
Underlying Dr. Niu’s research into the statistical applications of environmental studies is the belief that “environmental data analysis is an area that today is of critical importance in providing the basis of scientific understanding for setting wise public policy on environmental management on global and regional scales.” Researchers in the field of environmental data analysis are often motivated by the desire to solve problems in the real world. Developing new statistical theory to solve some of those problems creates numerous challenges, but the knowledge to be gained and the benefits to be reaped are incalculable.
Dr. Niu is currently teaching the graduate courses Statistics in Applications I and II. His other areas of research interest are time series analysis, spatial statistics, and linear and non-linear models.
Dr.’s Hollander and Sethuraman received nine percent salary increases
based on their 95-96 salaries
as part of the award.
The course which Zahn and Boroto designed is intended to help professionals learn how to deal more effectively with all aspects of their jobs: consultations, collaborations, supervision, and working with their bosses, the public, and members of Parliament. Zahn and Boroto take an innovative approach to their course by blending theory, group discussion, and feedback with video-taped role plays and coaching sessions that focus on improving one’s effectiveness in one-to-one and small group situations. Zahn’s wife, Andrea, also assists with the course.
Although attendees to the UK consultancy course occasionally enter feeling apprehensive about appearing on video, and skeptical as to what they may gain from the course, they frequently leave feeling very positive about how the experience will benefit them in the workplace and in life. The course has been so successful, in fact, that attendees now include trainers and information systems managers as well as statisticians.
The UK course has also benefited Zahn and Boroto in that the opportunity to work repeatedly with members of one organization over an extended period of time has helped them gather together work they have been doing over the past 18 years. Zahn and Boroto are currently using the information they have gathered to write a book on statistical consulting.
The University of Hong Kong’s assistance with the Marsaglia CD-ROM ensures that scholars will continue to have access to Dr. Marsaglia’s internationally famous work in random number generating and testing.
Doss, H., Huffer, F.W., and Lawson, K.L., Bayesian nonparametric estimation via Gibbs sampling for coherent systems with redundancy, Annals of Statistics, 25, 1109-1139, 1997.
Edward, F.C., Hosfield, E., Meeter, D.A., and Niu, X.-F., Tests for aggregation and size-based sample-unit selection when sample units vary in size, Ecology, 78 (4), 1238-1249, 1997: The paper estimates the degree of aggregation and tests for size-based sample-unit selection using generalized linear models.
Greenwood, P.E., McKeague, I.W., and Wefelmeyer, W., Outperforming the Gibbs sampler empirical estimator for nearest neighbor random fields, The Annals of Statistics, 24, 1433-1456, 1996.
Hollander, M., Boland, P.J., Joag-dev, K., and Kochar, S., Bivariate dependence properties of order statistics, Journal of Multivariate Analysis, 56, 75-89, 1996.
Hollander, M., McKeague, I.W., Li, G., and Yang, J., Nonparametric likelihood ratio confidence bands for quantile functions from incomplete survival data, Annals of Statistics, 24, 628-640, 1996.
Hollander, M., McKeague, I. W., and Yang, J., Likelihood based
confidence bands for survival functions, Journal of the American Statistical
Association, 92, 215-226, 1997. Abstract: Simultaneous
confidence bands are constructed using the nonparametric likelihood ratio
approach introduced by Thomas and Grunkemeier (1975). The boundaries
of the bands are contained within (0,1). A procedure almost equivalent
to a bias correction is developed, while the increase in coverage
precision is explained through an example and a simulation study. Likelihood ratio-based bands are developed for cumulative hazard functions.
Hollander, M., and Peña, E., Dynamic reliability models, Lifetime Data: Models in Reliability and Survival Analysis, (N.P. Jewell, A.C. Kimber, M.T. Lee and G.A. Whitmore, Eds.), 131-140, 1996, Kluwer Academic Publishers, Boston.
Hollander, M., and Peña, E., Reliability models and inference for series systems operating in different environments, Naval Logistics Research, 43, 1079-1108, 1996.
Huang, L.-S., Fan, J., Gijbels, I., and Hu, T.-C., An asymptotic study for variable bandwidth selection for a local polynomial regression, Statistica Sinica, 6, 113-127, 1996.
Huang, L.-S., and Leadbetter, M.R., On the statistics of exceedance
based environmental criteria, Proceedings of the A.C. Aitken Centenary
Conference (L. Kavalieris, F.C. Lam, L.A. Roberts, and J.A. Shanks,
Eds.), 173-181, 1996. Abstract: The current US “Ex-Ex”
criterion for ozone and two possible secondary criteria (“Area over threshold”
and “SUM06”) are used to illustrate a general class of compliance criteria
obtained as functions of excess values over threshold levels. Their
statistical properties are obtained from central limit theory for (stationary) array sums which gives Compound Poisson and normal approximations for such “exceedance statistics” above high and moderate threshold levels. The roles of level height, and the clustering of exceedances are discussed along with the distributional results obtained, in relatively non-technical terms. The Compound Poisson and normal type results given provide a basis for calculation of probabilities of correct compliance classification.
Livingston, R.J., Niu, X.-F.., Lewis, F.G., and Woodsum, G.C., Freshwater input to a gulf estuary: Long-term control of trophic organization, Ecological Applications, 7, 277-299, 1997: The paper studies the effects of reduction of freshwater flow on biological productivity in the Apalachicola Bay System.
McKeague, I.W., Variance reduction techniques for random fields, Mathematical Methods in Stochastic Simulation and Experimental Design (S. M. Ermakov, V. B. Melas, eds.), 202-208, 1996, St. Petersburg University Press.
McKeague, I.W., Aalen's additive risk model, Encyclopedia of Statistical Sciences, Update Volume 1 (S. Kotz and C. B. Read, Eds.), 1-6 ,1997, Wiley, New York.
McKeague, I.W., and Sun, Y., Transformations of Gaussian random fields to Brownian sheet and nonparametric change- point tests, Statistics and Probability Letters, 28, 311-319, 1996.
McKeague, I.W., and Sun, Y., Towards an omnibus distribution-free goodness-of-fit test for the Cox model, Statistica Sinica, 6, 579-588, 1996.
McKeague, I.W., and Zhang, M.-J., Fitting Cox's proportional hazard model using grouped survival data, Lifetime Data: Models in Reliability and Survival Analysis, 227-232, 1996.
Niu, X.-F., Nonlinear additive models for environmental time
series, with applications to ground-level ozone data analysis, Journal
of the American Statistical Association, 91, No 435, 1310-1321, 1996.
Abstract: A method for fitting additive models to an environmental
time series is introduced to characterize the nonlinear relationship between
the ozone concentration and meteorological variables as well as the serial
correlation of ozone data. The procedure derives maximum likelihood
estimates of noise parameters and implements back-fitting algorithms in regression analysis to estimate the model's unknown functions.
Niu, X.-F., Extreme value theory for a class of nonstationary time series with applications, The Annals of Applied Probability, 7 (2), 508-522, 1997.
Sethuraman, J., and Chaganty, N.R., Bahadur slopes for the t-statistic for a contaminated normal, Statistics and Probability Letters, 34, 245-250, 1997.
Sethuraman, J., and Chaganty, N.R., The large deviation principle for common statistical tests against a contaminated normal, Advances in Statistical Decision Theory and Methodology (in honor of Shanti S. Gupta), 1997.
Srivastava, A., Ergodic algorithms on special Euclidean groups for ATR, a chapter in Systems and Control in the 21st Century, in the Progress in Systems and Control series, 7, 1997.
Song, K.-S., and Muller, H.G., A set-indexed process in a two-region image, Stochastic Processes and Their Applications, 62, 87-101, 1996.
Song, K.-S., Two-stage change-point estimators in smooth regression models, Statistics and Probability Letters, 34, 323-335, 1997
Wu, Hulin, and Huffer, Fred W., Modeling the distribution of
plant species using the autologistic regression model, Environmental
and Ecological Statistics, 4, 49-64 , 1997.
Back to Contents
Doug Zahn attended the annual conference of the International
Alliance for Learning (IAL) in San Antonio in January 1997 where he and
his wife, Andrea, delivered an invited workshop on “Six aspects of teaching
and learning anything”.
Kai-Sheng Song presented an invited talk “On frequency estimation” at Eidgenossische Technische Hochschule, Zurich, Switzerland, March 1997.
Ian McKeague gave an invited talk at the special session on Spatial Stochastic Models at the American Mathematical Society Meeting, College Park, MD, April 1997
Fred Huffer was invited to McMaster University, Hamilton, Ontario, in June of 1997 to present a talk on “Approximating the distribution of the scan statistic using moments of the number of clumps”.
Anuj Srivastava gave a talk on “ A Bayesian approach to Automated Target Recognition and Tracking” at the Joint Research Conference on Statistics in Industry and Technology, New Brunswick, NJ, June 1997.
Duane Meeter gave an invited talk on “Empirical Bayes analysis of fractional factorial designs” to 3M statisticians in St. Paul, MN, July 1997.
Li-Shan Huang and Ian McKeague attended the annual meeting of the Institute of Mathematical Statistics in Park City, Utah in July 1997. Ian gave an invited talk on “Extracting information from random fields” at the meeting.
Jayaram Sethuraman attended the annual ASA meeting in Anaheim, CA in August 1997.
Doug Zahn and Dan Boroto (Psychology) presented an invited poster, “In statistical consulting there are no main effects, only interactions or consulting with clarity, certainty, and velocity”, at the ASA annual meeting in Anaheim, August 1997.
SUBRAMANIAN, SUNDARRAMAN, Department of Mathematics, University
of Maine, Estimation Under Censoring with Missing Failure Indicators (I.
YANG, JIE, John Stafford Trading, Chicago, Likelihood Ratio Based Confidence Interval in Survival Analysis (M. Hollander and I. McKeague, 1995).
CHEN, CHING-HSIANG, Department of Statistics, Tamkang University, Tamsui, Taiwan, R.O.C.,Identifying Influential
Effects in Factorial Experiments with Sixteen Runs: Empirical Bayes Approaches (D. Meeter and P. Lin, 1994)
LAWSON, KEVIN, Abbott Labs, Chicago, Bayesian Nonparametric Estimation Via Gibbs Sampling for Coherent Systems
With Redundancy (H. Doss and F. Huffer, 1994)
WU, HULIN, Frontier Science & Technology Research Foundation, Inc., Harvard University, Regression Models for Spatial Binary Data with Application to the Distribution of Plant Species (F. Huffer, 1993)
LIN, CHIEN-TAI, Department of Math, Tamkang University, Tamsui, Taiwan, R.O.C. The Computation of Probabilities
Which Involve Spacings, With Applications to the Scan Statistic (Huffer, 1993).
LEE, WEN-CHIUNG, Department of Statistics, Feng Chia University, Taichung, Taiwan, R.O.C. Generating Poisson and Binomial Random Variates (Marsaglia, 1993).
(1997), LEE, YOONJUNG, University of Wisconsin, at Madison, Madison, WI (1997), LEDDY, CHRISTINE, Barnett
Bank, Jacksonville, FL (1997), LOIZEAUX, MARC, Florida StateUniversity, Tallahassee, FL (1997), MCMICHAEL,
SYLWIA, (1997), MINTON, THOMAS, University of North Florida, Jacksonville, FL (1997), SCHONROCK,
HEATHER, Dynetics, Inc., Huntsville, AL (1997), SIMS, MELISSA, Tallahassee, FL (1997), WARRICK, SCOTT,
Applied Research Associates, Pace, FL (1997), TIGHIOUART, MOURAD, Florida State University (1997), BERCELI,
TERI, Pratt & Whitney, West Palm Beach, FL (1996), EAGAN, TAMMY, Department of Environmental Protection,
Tallahassee, FL (1996), HENTZ, JOSEPH, Abbot Laboratories, Abbott Park, Illinois (1996), MIRCHANDANI,
LAJWANTI (1996), PABEN, STEVEN, Bureau of Labor Statistics, Washington, DC (1996), STALLS, KENNETH,
Memphis Tennessee (1996), TURKOZ, IBRAHIM, Research and Data Services, Bloomingdale, IL (1996), YANG,
YIZHOU, Pittsburgh, PA (1996), ZHAO, FENG, Capital One, Richmond, VA (1996), LYLE, CHRISTOPHER T.,
United Airlines, Chicago, IL (1995), BOWLES (BURLESON) CATHERINE, University of Montana, Missoula, MT
(1994), ENSLEY, DAVID, Department of Corrections, Tallahassee, FL, Tallahassee, FL (1994), FOX, JOHN,
Environmental Protection Agency, Washington, DC (1994), LIU, MARIA, New England Research Institute, Watertown,
MA (1994), POTTER, MICHELLE, AT&T, New Jersey (1994), BURNETT, FRANK R., Science Applications
International Corp., Falls Church, VA (1993), DE POALO, TRACEY, Manugistics, Rockville, MD (1993), HALE,
ERIKA, Applied Logic Associates, Inc., Houston, TX (1993), LARSON, DIRK, Section of Biostatistics, Mayo Clinic, Rochester MN (1993)
Yongyuan was born in 1963 in the city of Huainan, Anhui province, People’s
Republic of China. He earned a bachelor’s degree in Mathematics in
1984 from Fuyang Normal University, and a Master’s degree in Mathematics
from Harbin Institute of Technology in 1987. In 1989 he came to the
United States and enrolled in FSU to pursue a Ph.D. in Mathematics.
As his studies progressed, Yongyuan developed a strong interest in mathematical statistics and stochastic processes and decided to transfer to the University of Georgia to pursue a degree in statistics. After he received his Master’s degree in Statistics in 1994, Yongyuan returned to Florida State and enrolled in the Ph.D. program in Statistics. He was working on his dissertation when he was diagnosed with colon cancer in August of 1996 and continued to work towards his Ph.D. until his death in April.
Yongyuan will be remembered as an intelligent, hard-working, and mature student whose enthusiasm and dedication towards his studies made a deep impression on all who knew him. To honor the spirit with which Yongyuan pursued his goals the Yongyuan and Anna Li Award Fund will be used by the department to award the graduate student who delivers the best talk in a contest of similar talks presented in the departmental colloquium series each year. Only graduate students who have been with the department for at least one year will be eligible to compete for the award each year.
The Yongyuan and Anna Li Award Fund was begun by a very generous donation from Yongyuan’s wife, Anna, and was furthered along through contributions from the department and from others. Yongyuan is survived by his wife and his son, Mao-Mao.