Christos Alexopoulos
Abstract
We extend and analyze a new class of estimators for the variance parameter of a steady-state simulation output process. These estimators are based on “folded” versions of the standardized time series (STS) of the process, and are analogous to the area and Cramér--von Mises estimators calculated from the original STS. In fact, one can apply the folding mechanism more than once to produce an entire class of estimators, all of which reuse the same underlying data stream. We show that these folded estimators share many of the same properties as their nonfolded counterparts, with the added bonus that they are often nearly independent of the nonfolded versions. In particular, we derive the asymptotic distributional properties of the various estimators as the run length increases, as well as their bias, variance, and mean squared error. We also study linear combinations of these estimators, and we show that such combinations yield estimators with lower variance than their constituents. Finally, we consider the consequences of batching, and we see that the batched versions of the new estimators compare favorably to benchmark estimators such as the nonoverlapping batch means estimator.
Supplemental Material
Available for Download
Online appendix to performance of folded variance estimators for simulation on article 11.
- }}Aktaran-Kalayci, T., Alexopoulos, C., Argon, N. T., Goldsman, D., and Wilson, J. R. 2007. Exact expected values of variance estimators in steady-state simulation. Naval Res. Logistics 54, 4, 397--410.Google Scholar
Cross Ref
- }}Alexopoulos, C., Argon, N. T., Goldsman, D., Steiger, N. M., Tokol, G., and Wilson, J. R. 2007a. Efficient computation of overlapping variance estimators for simulation. INFORMS J. Comput. 19, 3, 314--327. Google ScholarDigital Library
- }}Alexopoulos, C., Argon, N. T., Goldsman, D., Tokol, G., and Wilson, J. R. 2007b. Overlapping variance estimators for simulation. Oper. Res. 55, 6, 1090--1103. Google ScholarDigital Library
- }}Antonini, C. 2005. Folded variance estimators for stationary processes. Ph.D. dissertation, School of Industrial and Systems Engineering, Georgia Institute of Technology, Atlanta, GA.Google Scholar
- }}Antonini, C., Alexopoulos, C., Goldsman, D., and Wilson, J. R. 2007. Folded standardized time series area variance estimators for simulation. In Proceedings of the Winter Simulation Conference. S. G. Henderson et al. Eds., Institute of Electrical and Electronics Engineers, Piscataway, NJ, 455--462. Google ScholarDigital Library
- }}Antonini, C., Alexopoulos, C., Goldsman, D., and Wilson, J. R. 2009. Area variance estimators for simulation using folded standardized time series. IIE Trans. 41, 134--144.Google ScholarCross Ref
- }}Atkinson, K. 1988. An Introduction to Numerical Analysis. Wiley, New York.Google Scholar
- }}Billingsley, P. 1968. Convergence of Probability Measures. Wiley, New York.Google Scholar
- }}Box, G. E. P. 1954. Some theorems on quadratic forms applied in the study of analysis of variance problems, I. Effect of inequality of variance in the one-way classification. Ann. Math. Stat. 25, 290--302.Google ScholarCross Ref
- }}Calvin, J. M. 2007. Simulation output analysis using integrated paths. ACM Trans. Model. Comput. Simul. 17, 3, 13. Google ScholarDigital Library
- }}Calvin, J. M. and Nakayama, M. K. 2006. Permuted standardized time series estimators for steady-state simulation. Math. Oper. Res. 31, 2, 351--368. Google ScholarDigital Library
- }}Chien, C.-H., Goldsman, D., and Melamed, B. 1997. Large-sample results for batch means. Manage. Sci. 43, 1288--1295. Google ScholarDigital Library
- }}Conway, R. W. 1963. Some tactical problems in digital simulation. Manage. Sci. 10, 47--61.Google ScholarDigital Library
- }}Conway, R. W., Johnson, B. M., and Maxwell, M. L. 1959. Some problems of digital simulation. Manage. Sci. 6, 92--110.Google ScholarDigital Library
- }}Crane, M. A. and Lemoine, A. J. 1977. An Introduction to the Regenerative Method for Simulation Analysis. Springer, Berlin. Google ScholarDigital Library
- }}Fishman, G. S. 1973. Statistical analysis for queueing simulations. Manage. Sci. 20, 363--369.Google ScholarDigital Library
- }}Fishman, G. S. and Yarberry, L. S. 1997. An implementation of the batch means method. INFORMS J. Comput. 9, 296--310.Google ScholarDigital Library
- }}Foley, R. D. and Goldsman, D. 1999. Confidence intervals using orthonormally weighted standardized time series. ACM Trans. Model. Simul. 9, 297--325. Google ScholarDigital Library
- }}Glynn, P. W. and Iglehart, D. L. 1990. Simulation output analysis using standardized time series. Math. Oper. Res. 15, 1--16. Google ScholarDigital Library
- }}Glynn, P. W. and Whitt, W. 1991. Estimating the asymptotic variance with batch means. Oper. Res. Lett. 10, 431--435.Google ScholarDigital Library
- }}Goldsman, D., Kang, K., and Seila, A. F. 1999. Cramér--von Mises variance estimators for simulations. Oper. Res. 47, 299--309. Google ScholarDigital Library
- }}Goldsman, D. and Meketon, M. S. 1986. A comparison of several variance estimators. Tech. rep., School of Industrial and Systems Engineering, Georgia Institute of Technology, Atlanta, GA.Google Scholar
- }}Goldsman, D., Meketon, M. S., and Schruben, L. W. 1990. Properties of standardized time series weighted area variance estimators. Manage. Sci. 36, 602--612. Google ScholarDigital Library
- }}Goldsman, D. and Schruben, L. W. 1990. New confidence interval estimators using standardized time series. Manage. Sci. 36, 393--397. Google ScholarDigital Library
- }}Grenander, U. and Rosenblatt, M. 1957. Statistical Analysis of Stationary Time Series. Wiley, New York.Google Scholar
- }}Hald, A. 1952. Statistical Theory with Engineering Applications. Wiley, New York.Google Scholar
- }}Heidelberger, P. and Welch, P. D. 1981. A spectral method for confidence interval generation and run length control in simulations. Commun. ACM 24, 233--245. Google ScholarDigital Library
- }}Lada, E. K. and Wilson, J. R. 2006. A wavelet-based spectral procedure for steady-state simulation analysis. European J. Oper. Res. 174, 3, 1769--1801.Google ScholarCross Ref
- }}Law, A. M. 2007. Simulation Modeling and Analysis 4th Ed., McGraw-Hill, New York. Google ScholarDigital Library
- }}L'Ecuyer, P. 1999. Good parameters and implementations for combined multiple recursive random number generators. Oper. Res. 47, 159--164. Google ScholarDigital Library
- }}Meketon, M. S. and Schmeiser, B. W. 1984. Overlapping batch means: Something for nothing? In Proceedings of the Winter Simulation Conference. S. Sheppard et al. Eds., Institute of Electrical and Electronics Engineers, Piscataway, NJ, 227--230. Google ScholarDigital Library
- }}Satterthwaite, F. E. 1941. Synthesis of variance. Psychometrika 6, 309--316.Google ScholarCross Ref
- }}Schmeiser, B W. 1982. Batch size effects in the analysis of simulation output. Oper. Res. 30, 556--568.Google ScholarDigital Library
- }}Schruben, L. W. 1983. Confidence interval estimation using standardized time series. Oper. Res. 31, 1090--1108.Google ScholarDigital Library
- }}Shorack, G. R. and Wellner, J. A. 1986. Empirical Processes with Applications to Statistics. Wiley, New York.Google Scholar
- }}Song, W.-M. and Schmeiser, B. W. 1995. Optimal mean-squared-error batch sizes. Manage. Sci. 41, 110--123. Google ScholarCross Ref
- }}Steiger, N. M., Lada, E. K., Wilson, J. R., Joines, J. A., Alexopoulos, C., and Goldsman, D. 2005. ASAP3: A batch means procedure for steady-state simulation analysis. ACM Trans. Model. Comput. Simul. 15, 1, 39--73. Google ScholarDigital Library
- }}Steiger, N. M. and Wilson, J. R. 2001. Convergence properties of the batch-means method for simulation output analysis. INFORMS J. Comput.13, 4, 277--293. Google ScholarDigital Library
- }}Welch, P. D. 1987. On the relationship between batch means, overlapping batch means, and spectral estimation. In Proceedings of the Winter Simulation Conference. A. Thesen et al. Eds., Institute of Electrical and Electronics Engineers, Piscataway, NJ, 320--323. Google ScholarDigital Library
Index Terms
- Performance of folded variance estimators for simulation
Recommendations
Efficient Computation of Overlapping Variance Estimators for Simulation
For a steady-state simulation output process, we formulate efficient algorithms to compute certain estimators of the process variance parameter (i.e., the sum of covariances at all lags), where the estimators are derived in principle from overlapping ...
Properties of Standardized Time Series Weighted Area Variance Estimators
We wish to estimate the variance of the sample mean from a continuous-time stationary stochastic process. This article expands on the results of a technical note Goldsman and Schruben 1990 by using the theory of standardized time series to investigate ...
Optimal Mean-Squared-Error Batch Sizes
<P>When an estimator of the variance of the sample mean is parameterized by batch size, one approach for selecting batch size is to pursue the minimal mean squared error mse. We show that the convergence rate of the variance of the sample mean, and the ...
Comments