Stochastic dominance based comparison for system selection

We present two complementing selection procedures for comparing simulated systems based on the stochastic dominance relationship of a performance metric of interest. The decision maker specifies an output quantile set representing a section of the distribution of the metric, e.g., downside or upside risks or central tendencies, as the basis for comparison. The first procedure compares systems over the quantile set of interest by a first-order stochastic dominance criterion. The systems that are deemed nondominant in the first procedure could be compared by a weaker almost first-order stochastic dominance criterion in the second procedure. Numerical examples illustrate the capabilities of the proposed procedures.

[1]  F. Fred Choobineh,et al.  Do mean-based ranking and selection procedures consider systems' risk? , 2009, Proceedings of the 2009 Winter Simulation Conference (WSC).

[2]  W. David Kelton,et al.  Quantile and tolerance-interval estimation in simulation , 2006, Eur. J. Oper. Res..

[3]  H. Levy.  Stochastic Dominance: Investment Decision Making under Uncertainty , 2010 .

[4]  E. Paulson,et al.  Sequential Estimation and Closed Sequential Decision Procedures , 1964 .

[5]  Imrich Chlamtac,et al.  The P2 algorithm for dynamic calculation of quantiles and histograms without storing observations , 1985, CACM.

[6]  Moshe Leshno,et al.  Preferred by "All" and Preferred by "Most" Decision Makers: Almost Stochastic Dominance , 2002, Manag. Sci..

[7]  Murat Kulahci,et al.  Simulation-based cycle-time quantile estimation in manufacturing settings employing non-FIFO dispatching policies , 2009, J. Simulation.

[8]  Gerald T. Mackulak,et al.  Using quantiles in ranking and selection procedures , 2007, 2007 Winter Simulation Conference.

[9]  Herbert A. David,et al.  Order Statistics , 2011, International Encyclopedia of Statistical Science.

[10]  W. David Kelton,et al.  Estimating steady-state distributions via simulation-generated histograms , 2008, Comput. Oper. Res..

[11]  F. Fred Choobineh,et al.  A quantile-based approach to system selection , 2010, Eur. J. Oper. Res..

[12]  Milton Sobel,et al.  Nonparametric Procedures for Selecting the $t$ Population with the Largest $\alpha$-Quantiles , 1967 .

[13]  Chun-Hung Chen,et al.  Computing efforts allocation for ordinal optimization and discrete event simulation , 2000, IEEE Trans. Autom. Control..

[14]  Barry L. Nelson,et al.  A sequential procedure for neighborhood selection-of-the-best in optimization via simulation , 2006, Eur. J. Oper. Res..

[15]  Gerald T. Mackulak,et al.  Indirect cycle time quantile estimation using the Cornish–Fisher expansion , 2009 .

[16]  Frank E. Harrell,et al.  A new distribution-free quantile estimator , 1982 .

[17]  Stephen E. Chick,et al.  New Two-Stage and Sequential Procedures for Selecting the Best Simulated System , 2001, Oper. Res..

[18]  E. Parzen.  Nonparametric Statistical Data Modeling , 1979 .

[19]  Philip Heidelberger,et al.  Quantile Estimation in Dependent Sequences , 1984, Oper. Res..

[20]  RAJ JAIN,et al.  Histograms Without Storing Observations , 1999 .

[21]  Barry L. Nelson,et al.  A fully sequential procedure for indifference-zone selection in simulation , 2001, TOMC.

[22]  R. Parrish,et al.  Comparison of Quantile Estimators in Normal Sampling , 1990 .