Results 1  10
of
24
ASAP3: A batch means procedure for steadystate simulation analysis
 ACM Transactions on Modeling and Computer Simulation
, 2005
"... We introduce ASAP3, a refinement of the batch means algorithms ASAP and ASAP2, that delivers point and confidenceinterval estimators for the expected response of a steadystate simulation. ASAP3 is a sequential procedure designed to produce a confidenceinterval estimator that satisfies userspecif ..."
Abstract

Cited by 36 (23 self)
 Add to MetaCart
We introduce ASAP3, a refinement of the batch means algorithms ASAP and ASAP2, that delivers point and confidenceinterval estimators for the expected response of a steadystate simulation. ASAP3 is a sequential procedure designed to produce a confidenceinterval estimator that satisfies userspecified requirements on absolute or relative precision as well as coverage probability. ASAP3 operates as follows: the batch size is progressively increased until the batch means pass the ShapiroWilk test for multivariate normality; and then ASAP3 fits a firstorder autoregressive (AR(1)) time series model to the batch means. If necessary, the batch size is further increased until the autoregressive parameter in the AR(1) model does not significantly exceed 0.8. Next, ASAP3 computes the terms of an inverse CornishFisher expansion for the classical batch means tratio based on the AR(1) parameter estimates; and finally ASAP3 delivers a correlationadjusted confidence interval based on this expansion. Regarding not only conformance to the precision and coverageprobability requirements but also the mean and variance of the halflength of the delivered confidence interval, ASAP3 compared favorably to other batch means procedures (namely,
2005. Performance of a waveletbased spectral procedure for steadystate simulation analysis
 INFORMS Journal on Computing
, 2007
"... A summary and an analysis are given for an experimental performance evaluation of WASSP, an automated waveletbased spectral method for constructing an approximate confidence interval on the steadystate mean of a simulation output process such that the delivered confidence interval satisfies users ..."
Abstract

Cited by 11 (8 self)
 Add to MetaCart
A summary and an analysis are given for an experimental performance evaluation of WASSP, an automated waveletbased spectral method for constructing an approximate confidence interval on the steadystate mean of a simulation output process such that the delivered confidence interval satisfies userspecified requirements on absolute or relative precision as well as coverage probability. The experimentation involved three difficult test problems, each with an output process exhibiting some combination of the following characteristics: a long warmup period, a persistent autocorrelation structure, or a highly nonnormal marginal distribution. These problems were used to compare the performance of WASSP with that of the HeidelbergerWelch algorithm and ASAP3, two sequential procedures based respectively on the methods of spectral analysis and nonoverlapping batch means. Concerning efficiency (required sample sizes) and robustness against the statistical anomalies commonly encountered in simulation studies, WASSP outperformed the HeidelbergerWelch procedure and compared favorably with ASAP3. Key words: simulation, statistical analysis; spectral analysis; steadystate analysis; wavelet analysis
2011b. “Skart: A Skewness and AutoregressionAdjusted Batch Means Procedure for Simulation Analysis
 IIE Transactions
"... We discuss Skart, an automated batchmeans procedure for constructing a skewness and autoregressionadjusted confidence interval for the steadystate mean of a simulation output process. Skart is a sequential procedure designed to deliver a confidence interval that satisfies userspecified requir ..."
Abstract

Cited by 11 (8 self)
 Add to MetaCart
(Show Context)
We discuss Skart, an automated batchmeans procedure for constructing a skewness and autoregressionadjusted confidence interval for the steadystate mean of a simulation output process. Skart is a sequential procedure designed to deliver a confidence interval that satisfies userspecified requirements concerning not only coverage probability but also the absolute or relative precision provided by the halflength. Skart exploits separate adjustments to the halflength of the classical batchmeans confidence interval so as to account for the effects on the distribution of the underlying Student’s tstatistic that arise from nonnormality and autocorrelation of the batch means. Skart also delivers a point estimator for the steadystate mean that is approximately free of initialization bias. In an experimental performance evaluation involving a wide range of test processes, Skart compared favorably with other simulation analysis methods—namely, its predecessors ASAP3, WASSP, and SBatch as well as ABATCH, LBATCH, the HeidelbergerWelch procedure, and the LawCarson procedure.
NSkart: A Nonsequential Skewness and AutoregressionAdjusted BatchMeans Procedure for Simulation Analysis
"... Abstract—We discuss NSkart, a nonsequential procedure designed to deliver a confidence interval (CI) for the steadystate mean of a simulation output process when the user supplies a single simulationgenerated time series of arbitrary size and specifies the required coverage probability for a CI ba ..."
Abstract

Cited by 6 (3 self)
 Add to MetaCart
Abstract—We discuss NSkart, a nonsequential procedure designed to deliver a confidence interval (CI) for the steadystate mean of a simulation output process when the user supplies a single simulationgenerated time series of arbitrary size and specifies the required coverage probability for a CI based on that data set. NSkart is a variant of the method of batch means that exploits separate adjustments to the halflength of the CI so as to account for the effects on the distribution of the underlying Student’s tstatistic that arise from skewness (nonnormality) and autocorrelation of the batch means. If the sample size is sufficiently large, then NSkart delivers not only a CI but also a point estimator for the steadystate mean that is approximately free of initialization bias. In an experimental performance evaluation involving a wide range of test processes and sample sizes, NSkart exhibited close conformance to the userspecified CI coverage probabilities. Index Terms—autoregressive representation, confidence interval (CI), CornishFisher expansion, method of batch means, simulation, statistical analysis, steadystate analysis. I.
PERFORMANCE EVALUATION OF ASAP3 FOR STEADYSTATE OUTPUT ANALYSIS
"... An experimental performance evaluation of ASAP3 is presented, including several queueing systems with characteristics typically encountered in practical applications of steadystate simulation analysis procedures. Based on the method of nonoverlapping batch means, ASAP3 is a sequential procedure desi ..."
Abstract

Cited by 1 (1 self)
 Add to MetaCart
An experimental performance evaluation of ASAP3 is presented, including several queueing systems with characteristics typically encountered in practical applications of steadystate simulation analysis procedures. Based on the method of nonoverlapping batch means, ASAP3 is a sequential procedure designed to produce a confidenceinterval estimator of a steadystate mean response that satisfies userspecified precision and coverageprobability requirements. ASAP3 is compared with its predecessor ASAP and the batch means procedure of Law and Carson (LC) in the following test problems: (a) queue waiting times in the M/M/1/LIFO, M/H2/1, and M/M/1 queues with 80 % server utilization; and (b) response (sojourn) times in a central server model of a computer system. Regarding conformance to the given precision and coverageprobability requirements, ASAP3 compared favorably with the ASAP and LC procedures. Regarding the average sample sizes needed to satisfy progressively more stringent precision requirements, ASAP3’s efficiency was reasonable for the given test problems. 1
Folded Variance Estimators for Stationary Time Series Approved by:
, 2004
"... To my impatient mother, who never understood what was taking me so long. iii ACKNOWLEDGEMENTS I would like to begin by thanking my advisor, Dr. David Goldsman, my mentor, my friend, and the best person I have known. Your dedication and kindness to me and my profession went way beyond any of the res ..."
Abstract

Cited by 1 (0 self)
 Add to MetaCart
To my impatient mother, who never understood what was taking me so long. iii ACKNOWLEDGEMENTS I would like to begin by thanking my advisor, Dr. David Goldsman, my mentor, my friend, and the best person I have known. Your dedication and kindness to me and my profession went way beyond any of the responsibilities of an advisor. Thank you for being so approachable, dependable, and friendly even during the hardest times. You always believed in me even when I did not, and I will not let you down. It is a promise. Dr. Christos Alexopoulos, you had the hardest job, making my writing readable. Thank you for having checked every single line of my dissertation twenty thousand times, each time with the same fervor and devotion. I have learned a lot from your teachings, and I feel I am a much better professional right now. I always knew I had to polish myself, but you taught me how. It was an honor to have the opportunity to work with Dr. James Wilson, one of the
Steadystate simulation analysis using ASAP3
 In Proceedings of the 2004 Winter Simulation Conference, R.G.Ingalls
"... We discuss ASAP3, a refinement of the batch means algorithms ASAP and ASAP2. ASAP3 is a sequential procedure designed to produce a confidenceinterval estimator for the expected response of a steadystate simulation that satisfies userspecified precision and coverageprobability requirements. ASAP3 ..."
Abstract

Cited by 1 (0 self)
 Add to MetaCart
We discuss ASAP3, a refinement of the batch means algorithms ASAP and ASAP2. ASAP3 is a sequential procedure designed to produce a confidenceinterval estimator for the expected response of a steadystate simulation that satisfies userspecified precision and coverageprobability requirements. ASAP3 operates as follows: the batch size is increased until the batch means pass the ShapiroWilk test for multivariate normality; and then ASAP3 fits a firstorder autoregressive (AR(1)) time series model to the batch means. If necessary, the batch size is further increased until the autoregressive parameter in the AR(1) model does not significantly exceed 0.8. Next ASAP3 computes the terms of an inverse CornishFisher expansion for the classical batch means tratio based on the AR(1) parameter estimates; and finally ASAP3 delivers a correlationadjusted confidence interval based on this expansion. ASAP3 compared favorably with other batch means procedures (namely, ABATCH, ASAP, ASAP2, and LBATCH) in an extensive experimental performance evaluation. 1
Performance of Skart: A Skewness and AutoregressionAdjusted Batch Means Procedure for Simulation Analysis
, 2010
"... informs doi 10.1287/ijoc.1100.0401 ..."
CERTIFICATE OF APPROVAL
, 2013
"... To my parents. ii ACKNOWLEDGMENTS First and foremost I would like to thank my advisor Prof. Yong Chen. He has taught me a lot, both in research and course work. Not only instructed me about our research contents, he also spent efforts training me with good research habits. I really enjoy and appreci ..."
Abstract
 Add to MetaCart
To my parents. ii ACKNOWLEDGMENTS First and foremost I would like to thank my advisor Prof. Yong Chen. He has taught me a lot, both in research and course work. Not only instructed me about our research contents, he also spent efforts training me with good research habits. I really enjoy and appreciate my PhD years with him. I would like to thank my parents for being so supportive to me about everything since I was born. Also, I appreciate the encouragement and love from all my friends. iii In this thesis, there are generally three contributions to the Ranking and Selection problem in discreteevent simulation area. Ranking and selection is an important problem when people want to select single or multiple best designs from alternative pool. There are two different types in discreteevent simulation: terminating simulation
unknown title
"... We discuss Skart, an automated batchmeans procedure for constructing a skewness and autoregressionadjusted confidence interval (CI) for the steadystate mean of a simulation output process in either discrete time (i.e., observationbased statistics) or continuous time (i.e., timepersistent stati ..."
Abstract
 Add to MetaCart
(Show Context)
We discuss Skart, an automated batchmeans procedure for constructing a skewness and autoregressionadjusted confidence interval (CI) for the steadystate mean of a simulation output process in either discrete time (i.e., observationbased statistics) or continuous time (i.e., timepersistent statistics). Skart is a sequential procedure designed to deliver a CI that satisfies userspecified requirements concerning not only the CI’s coverage probability but also the absolute or relative precision provided by its halflength. Skart exploits separate adjustments to the halflength of the classical batchmeans CI so as to account for the effects on the distribution of the underlying Student’s tstatistic that arise from skewness (nonnormality) and autocorrelation of the batch means. The skewness adjustment is based on a modified CornishFisher expansion for the classical batchmeans Student’s tratio, and the autocorrelation adjustment is based on an autoregressive approximation to the batchmeans process for sufficiently large batch sizes. Skart also delivers a point estimator for the steadystate mean that is approximately free of initialization bias. The duration of the associated warmup period (i.e., the statistics clearing time) is based on iteratively applying von Neumann’s randomness test to spaced batch means with progressively increasing batch sizes and interbatch spacer sizes. In an experimental performance evaluation involving a wide range of test processes, Skart compared favorably with