Sequest: A Sequential Procedure for Estimating Quantiles in Steady-State Simulations

• 成果类型: Article
• 署名作者: Alexopoulos, Christos; Goldsman, David; Mokashi, Anup C.; Tien, Kal-Wen; Wilson, James R.
• 署名单位: University System of Georgia; Georgia Institute of Technology; SAS Institute Inc; Pennsylvania Commonwealth System of Higher Education (PCSHE); Pennsylvania State University; Pennsylvania State University - University Park; North Carolina State University
• 刊物名称: OPERATIONS RESEARCH
• ISSN/ISSBN: 0030-364X
• DOI: 10.1287/opre.2018:1829
• 发表日期: 2019
• 页码: 1162-1183
• 关键词: batch-means procedure; stable stochastic-systems; mean-squared-error; bahadur representation; sample quantiles; skewness; performance; skart; distributions; ratio

摘要

Sequest is an automated sequential simulation-analysis procedure designed to provide improved point and confidence-interval (CI) estimators for a designated steady-state quantile by the use of batching and sectioning. The procedure incorporates techniques to do the following: (i) reduce the absolute bias of the sectioning-based point estimator caused by atypical simulation initialization or an inadequate sample size, (ii) adjust the batching-based CI half-length to compensate for correlation or skewness of the batch quantile estimators used in computing the CI half-length, and (iii) obtain a sample size sufficient to yield a CI satisfying user-specified requirements on the CI's coverage probability and its absolute or relative precision. In a set of difficult test problems, Sequest compared favorably with its competitors, requiring smaller average sample sizes while delivering better CI conformance to the given coverage and precision requirements.