Professor  Emeritus
Department of Operations Research
University of North Carolina at Chapel Hill

134 Hanes Hall
UNC Campus 
Telephone:            (919) 962-3835
Departmental tel.:  (919) 962-1275
FAX:                    (919) 962-0391 
E-mail: gfish@email.unc.edu
Mailing Address:
    CB #3260, UNC
    Chapel Hill, NC 27599-3260

Discrete-event simulation
Monte Carlo sampling 
Design and analysis of large-scale simulation experiments

• • A First Course in Monte Carlo, Thomson Brooks/Cole, 2005. For errata, see fcmc_errata.pdf.

• • Discrete-Event Simulation: Modeling, Programming, and Analysis, New York, Springer-Verlag, 2001. For errata, see des_errata.pdf .

• •  Monte Carlo: Concepts, Algorithms, and  Applications, New York, Springer-Verlag, 1996.
    
  • An analysis of  Swendsen-Wang and related sampling methods, J. Roy. Statist. Soc., B, 61, Part 3, (1999) 623-641.

• • Best- and worst-case variances when bounds are available for the distribution function, Computational Statistics and Data Analysis, 29, (1998), 35-53. 

  LABATCH.2  is a collection of programs for statistically analyzing sample path data on a strictly stationary stochastic process. For each sample path in its input, it computes a sample mean and an asymptotically valid confidence interval for the population mean. It also displays a sequence of estimates of the asymptotic variance of the sample mean that allows an assessment of the quality of the final estimate of this quantity and, consequently, an assessment of the validity of the confidence interval. fcmc contains a collection of programs that can facilitate Mont Carlo sampling.   In particular, for  the pseudorandom number generator
known as the Mersenne twister, MT19937, a C-language program, rng_afm.c, provides a means for loading seeds at the beginning of a run
and saving the final numbers in the sub-sequence at the end of the run.
 This allows for non-overlapping sub-sequences on successive runs.

seed_file contains seeds to initialize rng_afm.c.

ct.hm.c estimates the number of two-way contingency tables with given  row sums and column sums.

ct.chisq.c estimates the number of contingency tables with chi-squared statistics no larger than a given value.

Advancing the Frontiers of Simulation: A Festschrift in Honor of George Samuel Fishman, C. Alexopoulos, D. Goldsman, J. R. Wilson, eds, Springer 2009.

Lanchester Prize (INFORMS) 
for the 1996 outstanding publication in Operations Research and Management Sciences in English.INFORMS College on Simulation Award for the 1997 outstanding publication in simulation.

Distinguished Service Award (1990)
INFORMS College on Simulation

Ph.D. (1970) in Biostatistics UCLA 

Analyst
The RAND Corporation 
1962-70

Associate Professor 
Administrative Sciences Department Yale University
1970-74

Professor
Operations Research Department University of North Carolina 
Chapel Hill, NC
1974-present
Complete_Curriculum_Vitae