Marvin K. Nakayama
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Marvin K. Nakayama has a total of 19 co-authors. Here are the top co-authors by number of shared publications:
- Bruno Tuffin
INRIA
4 shared publications - Jose Blanchet
Department of Management Science & Engineering Stanford University
2 shared publications - Juan Li
Department of Industrial Engineering and Operations Research, Columbia University, New York, New York 10027
2 shared publications - Yajuan Li
Department of Computer Science, New Jersey Institute of Technology, Newark, New Jersey 07102
2 shared publications - Zachary T. Kaplan
Department of Computer Science, New Jersey Institute of Technology, Newark, New Jersey 07102
2 shared publications - Alexandros V. Gerbessiotis
CS Department, New Jersey Institute of Technology, Newark, NJ 07102, USA
1 shared publication
Latest Publications
- Open Access
- 2 October 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Modeling and Computer Simulation
Efficiency of Estimating Smooth Functions of Parameters in Rare-Event Contexts
Consider an estimand that is a known smooth function of an unknown vector of parameters, such as means, probabilities, quantiles, etc. Some but not necessarily all of the parameters relate to rare events, which typically make them difficult to estimate with naive Monte Carlo, so variance-reduction techniques must be applied to obtain efficient estimators of those parameters. We develop conditions ensuring that the overall estimand can be efficiently estimated, providing both sufficient and necessary-and-sufficient conditions. The paper illustrates the applicability of the theory through many examples and numerical results.
- 2 October 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Modeling and Computer Simulation
Open AccessEfficiency of Estimating Smooth Functions of Parameters in Rare-Event Contexts
Consider an estimand that is a known smooth function of an unknown vector of parameters, such as means, probabilities, quantiles, etc. Some but not necessarily all of the parameters relate to rare events, which typically make them difficult to estimate with naive Monte Carlo, so variance-reduction techniques must be applied to obtain efficient estimators of those parameters. We develop conditions ensuring that the overall estimand can be efficiently estimated, providing both sufficient and necessary-and-sufficient conditions. The paper illustrates the applicability of the theory through many examples and numerical results.
- Open Access
- 16 May 2024
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Modeling and Computer Simulation
- Vol. 34 (3) , 1-38
Sufficient Conditions for Central Limit Theorems and Confidence Intervals for Randomized Quasi-Monte Carlo Methods
Randomized quasi-Monte Carlo methods have been introduced with the main purpose of yielding a computable measure of error for quasi-Monte Carlo approximations through the implicit application of a central limit theorem over independent randomizations. But to increase precision for a given computational budget, the number of independent randomizations is usually set to a small value so that a large number of points are used from each randomized low-discrepancy sequence to benefit from the fast convergence rate of quasi-Monte Carlo. While a central limit theorem has been previously established for a specific but computationally expensive type of randomization, it is also known in general that fixing the number of randomizations and increasing the length of the sequence used for quasi-Monte Carlo can lead to a non-Gaussian limiting distribution. This paper presents sufficient conditions on the relative growth rates of the number of randomizations and the quasi-Monte Carlo sequence length to ensure a central limit theorem and also an asymptotically valid confidence interval. We obtain several results based on the Lindeberg condition for triangular arrays and expressed in terms of the regularity of the integrand and the convergence speed of the quasi-Monte Carlo method. We also analyze the resulting estimator’s convergence rate.
- 16 May 2024
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Modeling and Computer Simulation
- Vol. 34 (3) , 1-38
Open AccessSufficient Conditions for Central Limit Theorems and Confidence Intervals for Randomized Quasi-Monte Carlo Methods
Randomized quasi-Monte Carlo methods have been introduced with the main purpose of yielding a computable measure of error for quasi-Monte Carlo approximations through the implicit application of a central limit theorem over independent randomizations. But to increase precision for a given computational budget, the number of independent randomizations is usually set to a small value so that a large number of points are used from each randomized low-discrepancy sequence to benefit from the fast convergence rate of quasi-Monte Carlo. While a central limit theorem has been previously established for a specific but computationally expensive type of randomization, it is also known in general that fixing the number of randomizations and increasing the length of the sequence used for quasi-Monte Carlo can lead to a non-Gaussian limiting distribution. This paper presents sufficient conditions on the relative growth rates of the number of randomizations and the quasi-Monte Carlo sequence length to ensure a central limit theorem and also an asymptotically valid confidence interval. We obtain several results based on the Lindeberg condition for triangular arrays and expressed in terms of the regularity of the integrand and the convergence speed of the quasi-Monte Carlo method. We also analyze the resulting estimator’s convergence rate.
- Open Access
- 30 April 2024
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Modeling and Computer Simulation
- Vol. 34 (2) , 1-3
Introduction to the Special Issue for INFORMS Simulation Society (I-Sim) Workshop, 2021
- 30 April 2024
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Modeling and Computer Simulation
- Vol. 34 (2) , 1-3
Open AccessIntroduction to the Special Issue for INFORMS Simulation Society (I-Sim) Workshop, 2021
- 1 January 2024
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in INFORMS Journal on Computing
- Vol. 36 (1) , 266-284
Monte Carlo Methods for Economic Capital
Economic capital (EC) is a risk measure used by financial firms to specify capital levels to protect (with high probability) against large unforeseen losses. Defined as the difference between an (extreme) quantile and the mean of the loss distribution, the EC is often estimated via Monte Carlo methods. Although simple random sampling (SRS) may be effective in estimating the mean, it can be inefficient for the extreme quantile in the EC. Applying importance sampling (IS) may lead to an efficient quantile estimator but can do poorly for the mean. Measure-specific IS (MSIS) instead uses IS to estimate only the quantile, and the mean is independently handled via SRS. We analyze large-sample properties of EC estimators obtained via SRS only, IS only, MSIS, IS using a defensive mixture, and a double estimator using both SRS and IS to estimate both the quantile and the mean, establishing Bahadur-type representations for the EC estimators and proving they obey central limit theorems. We provide asymptotic theory comparing the estimators when the loss is the sum of a large number of independent and identically distributed random variables. Numerical and simulation results, including for a large portfolio credit risk model with dependent obligors, complement the theory.History: Accepted by Bruno Tuffin, Area Editor for Simulation.Funding: This work was supported by the National Science Foundation [Grant CMMI-1537322].Supplemental Material: The software that supports the findings of this study is available within the paper and its Supplemental Information ( https://pubsonline.informs.org/doi/suppl/10.1287/ijoc.2021.0261 ) as well as from the IJOC GitHub software repository ( https://github.com/INFORMSJoC/2021.0261 ). The complete IJOC Software and Data Repository is available at https://informsjoc.github.io/ .
- 1 January 2024
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in INFORMS Journal on Computing
- Vol. 36 (1) , 266-284
Monte Carlo Methods for Economic Capital
Economic capital (EC) is a risk measure used by financial firms to specify capital levels to protect (with high probability) against large unforeseen losses. Defined as the difference between an (extreme) quantile and the mean of the loss distribution, the EC is often estimated via Monte Carlo methods. Although simple random sampling (SRS) may be effective in estimating the mean, it can be inefficient for the extreme quantile in the EC. Applying importance sampling (IS) may lead to an efficient quantile estimator but can do poorly for the mean. Measure-specific IS (MSIS) instead uses IS to estimate only the quantile, and the mean is independently handled via SRS. We analyze large-sample properties of EC estimators obtained via SRS only, IS only, MSIS, IS using a defensive mixture, and a double estimator using both SRS and IS to estimate both the quantile and the mean, establishing Bahadur-type representations for the EC estimators and proving they obey central limit theorems. We provide asymptotic theory comparing the estimators when the loss is the sum of a large number of independent and identically distributed random variables. Numerical and simulation results, including for a large portfolio credit risk model with dependent obligors, complement the theory.History: Accepted by Bruno Tuffin, Area Editor for Simulation.Funding: This work was supported by the National Science Foundation [Grant CMMI-1537322].Supplemental Material: The software that supports the findings of this study is available within the paper and its Supplemental Information ( https://pubsonline.informs.org/doi/suppl/10.1287/ijoc.2021.0261 ) as well as from the IJOC GitHub software repository ( https://github.com/INFORMSJoC/2021.0261 ). The complete IJOC Software and Data Repository is available at https://informsjoc.github.io/ .
- 8 September 2023
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in INFORMS Journal on Computing
Code and Data Repository for Monte Carlo Methods for Economic Capital
This repository contains following folders: data, results, src. data: includes data needed for the PCRM (portfolio credit risk model) code. results: contains figure and table results for both PCRM and i.i.d. sum model. src: includes two folders. The folder "PCRM" contains the codes for portfolio credit risk model in the paper, which are written in python. The other folder "IID" contains codes for i.i.d. sum model, which are written in matlab.
- 8 September 2023
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in INFORMS Journal on Computing
Code and Data Repository for Monte Carlo Methods for Economic Capital
This repository contains following folders: data, results, src. data: includes data needed for the PCRM (portfolio credit risk model) code. results: contains figure and table results for both PCRM and i.i.d. sum model. src: includes two folders. The folder "PCRM" contains the codes for portfolio credit risk model in the paper, which are written in python. The other folder "IID" contains codes for i.i.d. sum model, which are written in matlab.
- 1 December 2022
- Published by Springer Nature
- p. 333-351
Array-RQMC to Speed up the Simulation for Estimating the Hitting-Time Distribution to a Rare Set of a Regenerative System
- 1 December 2022
- Published by Springer Nature
- p. 333-351
Array-RQMC to Speed up the Simulation for Estimating the Hitting-Time Distribution to a Rare Set of a Regenerative System
- Open Access
Sufficient Conditions for Central Limit Theorems and Confidence Intervals for Randomized Quasi-Monte Carlo Methods
Randomized quasi-Monte Carlo methods have been introduced with the main purpose of yielding a computable measure of error for quasi-Monte Carlo approximations through the implicit application of a central limit theorem over independent randomizations. But to increase precision for a given computational budget, the number of independent randomizations is usually set to a small value so that a large number of points are used from each randomized low-discrepancy sequence to benefit from the fast convergence rate of quasi-Monte Carlo. While a central limit theorem has been previously established for a specific but impractical type of randomization, it is also known in general that fixing the number of randomizations and increasing the length of the sequence used for quasi-Monte Carlo can lead to a non-Gaussian limiting distribution. This paper presents sufficient conditions on the relative growth rates of the number of randomizations and the quasi-Monte Carlo sequence length to ensure a central limit theorem and also an asymptotically valid confidence interval. We obtain several results based on the Lindeberg condition and expressed in terms of the regularity of the integrand and the convergence speed of the quasi-Monte Carlo method. We also analyze the resulting estimator's convergence rate.
Open AccessSufficient Conditions for Central Limit Theorems and Confidence Intervals for Randomized Quasi-Monte Carlo Methods
Randomized quasi-Monte Carlo methods have been introduced with the main purpose of yielding a computable measure of error for quasi-Monte Carlo approximations through the implicit application of a central limit theorem over independent randomizations. But to increase precision for a given computational budget, the number of independent randomizations is usually set to a small value so that a large number of points are used from each randomized low-discrepancy sequence to benefit from the fast convergence rate of quasi-Monte Carlo. While a central limit theorem has been previously established for a specific but impractical type of randomization, it is also known in general that fixing the number of randomizations and increasing the length of the sequence used for quasi-Monte Carlo can lead to a non-Gaussian limiting distribution. This paper presents sufficient conditions on the relative growth rates of the number of randomizations and the quasi-Monte Carlo sequence length to ensure a central limit theorem and also an asymptotically valid confidence interval. We obtain several results based on the Lindeberg condition and expressed in terms of the regularity of the integrand and the convergence speed of the quasi-Monte Carlo method. We also analyze the resulting estimator's convergence rate.
- 1 September 2019
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Operations Research
- Vol. 67 (5) , 1383-1396
Rare-Event Simulation for Distribution Networks
Catastrophic events can dramatically disrupt distribution networks and supply chains; think, for example, about the threat of a hurricane, which causes significant spikes in demands for gasoline, water, and other products. What are the weak links in such a supply network, or what is the most likely way in which such a network can be disrupted? In “Rare Event Simulation for Distribution Networks,” J. Blanchet, J. Li, and M. K. Nakayama devise optimal simulation algorithms for estimating the occurrence of disruptive events in complex distribution networks. This development is noteworthy because it can be used to mitigate the impact in populations that are susceptible to these types of events, which can incur tremendous costs to society.
- 1 September 2019
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Operations Research
- Vol. 67 (5) , 1383-1396
Rare-Event Simulation for Distribution Networks
Catastrophic events can dramatically disrupt distribution networks and supply chains; think, for example, about the threat of a hurricane, which causes significant spikes in demands for gasoline, water, and other products. What are the weak links in such a supply network, or what is the most likely way in which such a network can be disrupted? In “Rare Event Simulation for Distribution Networks,” J. Blanchet, J. Li, and M. K. Nakayama devise optimal simulation algorithms for estimating the occurrence of disruptive events in complex distribution networks. This development is noteworthy because it can be used to mitigate the impact in populations that are susceptible to these types of events, which can incur tremendous costs to society.
- 1 December 2017
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Operations Research
- Vol. 65 (6) , 1678-1695
Quantile Estimation with Latin Hypercube Sampling
Quantiles are often used to measure risk of stochastic systems. We examine quantile estimators obtained using simulation with Latin hypercube sampling (LHS), a variance-reduction technique that efficiently extends stratified sampling to higher dimensions and produces negatively correlated outputs. We consider single-sample LHS (ssLHS), which minimizes the variance that can be obtained from LHS, and also replicated LHS (rLHS). We develop a consistent estimator of the asymptotic variance of the ssLHS quantile estimator’s central limit theorem, enabling us to provide the first confidence interval (CI) for a quantile when applying ssLHS. For rLHS, we construct CIs using batching and sectioning. On average, our rLHS CIs are shorter than previous rLHS CIs and only slightly wider than the ssLHS CI. We establish the asymptotic validity of the CIs by first proving that the quantile estimators satisfy Bahadur representations, which show that the quantile estimators can be approximated by linear transformations of estimators of the cumulative distribution function. We present numerical results comparing the various CIs.
- 1 December 2017
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Operations Research
- Vol. 65 (6) , 1678-1695
Quantile Estimation with Latin Hypercube Sampling
Quantiles are often used to measure risk of stochastic systems. We examine quantile estimators obtained using simulation with Latin hypercube sampling (LHS), a variance-reduction technique that efficiently extends stratified sampling to higher dimensions and produces negatively correlated outputs. We consider single-sample LHS (ssLHS), which minimizes the variance that can be obtained from LHS, and also replicated LHS (rLHS). We develop a consistent estimator of the asymptotic variance of the ssLHS quantile estimator’s central limit theorem, enabling us to provide the first confidence interval (CI) for a quantile when applying ssLHS. For rLHS, we construct CIs using batching and sectioning. On average, our rLHS CIs are shorter than previous rLHS CIs and only slightly wider than the ssLHS CI. We establish the asymptotic validity of the CIs by first proving that the quantile estimators satisfy Bahadur representations, which show that the quantile estimators can be approximated by linear transformations of estimators of the cumulative distribution function. We present numerical results comparing the various CIs.