Using subsystem linear regression metamodels in stochastic simulation

This article explores the use of metamodels as simulation building blocks. The metamodel replaces a part of the simulation model with a mathematical function that mimics the input–output behavior of that part, with respect to some measure of interest to the designer. The integration of metamodels as components of the simulation model simplifies the model and reduces the simulation time. Such use of the metamodels also gives the designer a better understanding of the behavior of those parts of the model, making the simulation model as a whole more intelligible. The metamodel-based simulation model building process is examined, step by step, and the designer options are explored. This process includes the identification of the metamodel candidates and the construction of the metamodels themselves. The assessment of the proposed approach includes the evaluation of the integration effort of the metamodel into the metamodel-based simulation model, and the accuracy of the output data when compared to the original system.     

Using DEA to assess NATO burden-sharing

As the member making many of the most visible contributions to NATO, the USA has often claimed that they shoulder the heaviest ‘burden’ in maintaining the objectives of the alliance. This claim has been backed by research which has concentrated on contributions such as defence expenditure and benefits such as protection from external threats. However, modern alliances entail the exchange of multiple forms of alliance benefits and liabilities and therefore the study of ‘burden-sharing’ in these alliances is a more complicated accounting problem than has been acknowledged by previous research. In this paper, burden-sharing is studied using Data Envelopment Analysis (DEA). This is a novel application of DEA. Instead of calculating relative ‘efficiencies’, the analysis produces a ‘net-burden index’ for each member nation of NATO. The results of the analysis indicate that whilst the USA shoulders a heavy burden, some other member nations, including Canada and Spain, may in fact shoulder a heavier burden.     

Discretization and simulation of stochastic differential equations

We discuss both pathwise and mean-square convergence of several approximation schemes to stochastic differential equations. We then estimate the corresponding speeds of convergence, the error being either the mean square error or the error induced by the approximation on the value of the expectation of a functional of the solution. We finally give and comment on a few comparative simulation results.     

Error and convergence of two numerical schemes for stochastic differential equations

We examine the convergence and error rate of two stochastic numerical schemes using the method of proof used by G. N. Mil'shtein 1 . © 2006 Wiley Periodicals, Inc. Numer Methods Partial Differential Eq, 2006.     

Using linear programming to analyze and optimize stochastic flow lines

This paper presents a linear programming approach to analyze and optimize flow lines with limited buffer capacities and stochastic processing times. The basic idea is to solve a huge but simple linear program that models an entire simulation run of a multi-stage production process in discrete time, to determine a production rate estimate. As our methodology is purely numerical, it offers the full modeling flexibility of stochastic simulation with respect to the probability distribution of processing times. However, unlike discrete-event simulation models, it also offers the optimization power of linear programming and hence allows us to solve buffer allocation problems. We show under which conditions our method works well by comparing its results to exact values for two-machine models and approximate simulation results for longer lines.     

Two-stage stochastic equilibrium problems with equilibrium constraints: modeling and numerical schemes

This article presents a two-stage stochastic equilibrium problem with equilibrium constraints (SEPEC) model. Some source problems which motivate the model are discussed. Monte Carlo sampling method is applied to solve the SEPEC. Convergence analysis on the statistical estimators of Nash equilibria and Nash stationary points are presented.     

Euler schemes and large deviations for stochastic Volterra equations with singular kernels

In this paper, an Euler type approximation is constructed for stochastic Volterra equation with singular kernels, which provides an algorithm for numerical calculation. Then, the large deviation estimates of small perturbation to equations of this type are obtained. We finally apply them to SDEs with the kernel of fractional Brownian motion with Hurst parameter H∈(0,1).     

Using spatial games to model dynamic evolutionary systems

Modelling the dynamics of evolutionary competing species on a physical grid is a challenging modelling problem. This paper presents a novel modelling approach for synthesizing evolutionary dynamics of competing species using a spatial game perspective. This modelling approach describes the movement of players (‘species’ in our context) across a lattice. The model is based on a payoff function which controls the move likelihood and direction of the players (‘predators’ and ‘preys’). Using simulated results, the paper provides a comparison between the spatial game model and an existing predator-prey dynamic model. Finally, a case study is performed to illustrate the application of this formalism and validate the model.     

Survival and coexistence in stochastic spatial Lotka–Volterra models

A spatially explicit, stochastic Lotka–Volterra model was introduced by Neuhauser and Pacala in Neuhauser and Pacala (Ann. Appl. Probab. 9, 1226–1259, 1999). A low density limit theorem for this process was proved by the authors in Cox and Perkins (Ann. Probab. 33, 904–947, 2005), showing that certain generalized rescaled Lotka–Volterra models converge to super-Brownian motion with drift. Here we use this convergence result to extend what is known about the parameter regions for the Lotka–Volterra process where (i) survival of one type holds, and (ii) coexistence holds. Supported in part by an NSERC Research grant.     

Risk measurement and management of defined benefit pension schemes: a stochastic approach

The traditional actuarial valuation for defined benefit pensionschemes operates on the basis of a set of deterministic calculationscombined with actuarial judgment. It has played an importantrole in guiding decision-making as far as the level of fundingis concerned. The paper argues that stochastic methods can addvalue in certain crucial areas, in particular the financialrisk management of such schemes. The traditional approach torisk is to incorporate margins in the valuation assumptions;however, a stochastic approach allows the user to evaluate specificand quantifiable risk and performance measures in respect ofalternative funding and investment strategies. The paper introducesa framework that measures the risks inherent in asset allocationand contribution rate decisions, allowing decisions to be madeon a more informed basis. In doing this, we suggest and applysome potential risk and performance measures. This frameworkprovides the means to explore the trade-offs involved in possiblecontribution and asset allocation decisions and leads to decisionstrategies that are expected to give improved outcomes for thesame level of risk. A realistic case study is used to illustratethe properties of the methodology and how it might be used.     

Weak local linear discretizations for stochastic differential equations: Convergence and numerical schemes

Weak local linear (WLL) discretizations are playing an increasing role in the construction of effective numerical integrators and inference methods for stochastic differential equations (SDEs) with additive noise. However, due to limitations in the existing numerical implementations of WLL discretizations, the resulting integrators and inference methods have either been restricted to particular classes of autonomous SDEs or showed low computational efficiency. Another limitation is the absence of a systematic theoretical study of the rate of convergence of the WLL discretizations and numerical integratos. This task is the main purpose of the present paper. A second goal is introducing a new WLL scheme that overcomes the numerical limitations mentioned above. Additionally, a comparative analysis between the new WLL scheme and some conventional weak integrators is also presented.     

Structured construction and simulation of nondeterministic stochastic activity networks

In this work we deal with nondeterministic stochastic activity networks (NDSANs). Their stochastic character results from activity durations, which are given by nonnegative continuous random variables. The nondeterministic behavior of an NDSAN is a consequence of its variable topology, based on two additional features. First, by associating choice probabilities with the immediate successors of an activity, some branches of execution are not always taken. Second, by allowing iterated executions of a group of activities according to predetermined probabilities, the number of times an activity is to be executed is not determined a priori. These properties lead to a wide variety of activity networks, capable of modelling many real situations in process engineering and project management. We describe a simple, recursively structured construction of NDSANs, which both provides a coherent syntactic mechanism to incorporate the two abovementioned nondeterminism features and allows the analytic formulation of completion time. This construction also directly gives rise to a recursive simulation algorithm for NDSANs, whose repeated execution produces an estimate of the probability distribution of the completion time of the network. We also report on real-world case studies, using the Komolgorov–Smirnov statistic for validation.     

Using simulation to improve the blood supply chain

This case study is concerned with analysing policies for managing the blood inventory system in a typical UK hospital supplied by a regional blood centre. The objective of the project is to improve procedures and outcomes by modelling the entire supply chain for that hospital, from donor to recipient. The supply chain of blood products is broken down into material flows and information flows. Discrete-event simulation is used to determine ordering policies leading to reductions in shortages and wastage, increased service levels, improved safety procedures and reduced costs, by employing better system coordination. In this paper we describe the model and present results for a representative medium-sized hospital. The model can be used by both the National Blood Service and by hospital managers as a decision support tool to investigate different procedures and policies.     

Using genetic algorithm to solve a new multi-period stochastic optimization model

This paper presents a new asset allocation model based on the CVaR risk measure and transaction costs. Institutional investors manage their strategic asset mix over time to achieve favorable returns subject to various uncertainties, policy and legal constraints, and other requirements. One may use a multi-period portfolio optimization model in order to determine an optimal asset mix. Recently, an alternative stochastic programming model with simulated paths was proposed by Hibiki [N. Hibiki, A hybrid simulation/tree multi-period stochastic programming model for optimal asset allocation, in: H. Takahashi, (Ed.) The Japanese Association of Financial Econometrics and Engineering, JAFFE Journal (2001) 89-119 (in Japanese); N. Hibiki A hybrid simulation/tree stochastic optimization model for dynamic asset allocation, in: B. Scherer (Ed.), Asset and Liability Management Tools: A Handbook for Best Practice, Risk Books, 2003, pp. 269-294], which was called a hybrid model. However, the transaction costs weren’t considered in that paper. In this paper, we improve Hibiki’s model in the following aspects: (1) The risk measure CVaR is introduced to control the wealth loss risk while maximizing the expected utility; (2) Typical market imperfections such as short sale constraints, proportional transaction costs are considered simultaneously. (3) Applying a genetic algorithm to solve the resulting model is discussed in detail. Numerical results show the suitability and feasibility of our methodology.     

Deterministic solution and stochastic simulation of a simple production-inventory model

Summary In the literature to date there have only been case study applications of corporate models with a structure expressed in first order coupled differential or difference equations. The paper deals with the optimal analytic and deterministic solution of a corporate submodel using the maximum principle. A stochastic extension using Monte-Carlo simulation together with variance reduction techniques is also discussed.

---

Zusammenfassung Bis jetzt wurden Unternehmensmodelle, deren Struktur sich durch gekoppelte Differentialoder Differenzgleichungen darstellen ließ, nur in Fallstudien eingesetzt. Die vorliegende Arbeit zeigt, wie eine optimale deterministische Lösung eines einfachen Untermodells durch Anwendung des Maximum-Prinzips gefunden werden kann. Für zufällig streuende Modellparameter wird eine optimale Lösung unter Verwendung der Monte-Carlo-Methode und varianzreduzierenden Methoden empirisch hergeleitet.

     

Asymptotic distributions of M-estimators in a spatial regression model under some fixed and stochastic spatial sampling designs

In this paper, we consider M-estimators of the regression parameter in a spatial multiple linear regression model. We establish consistency and asymptotic normality of the M-estimators when the data-sites are generated by a class of deterministic as well as a class of stochastic spatial sampling schemes. Under the deterministic sampling schemes, the data-sites are located on a regular grid but may have aninfill component. On the other hand, under the stochastic sampling schemes, locations of the data-sites are given by the realizations of a collection of independent random vectors and thus, are irregularly spaced. It is shown that scaling constants of different orders are needed for asymptotic normality under different spatial sampling schemes considered here. Further, in the stochastic case, the asymptotic covariance matrix is shown to depend on the spatial sampling density associated with the stochastic design. Results are established for M-estimators corresponding to certain non-smooth score functions including Huber’s ψ-function and the sign functions (corresponding to the sample quantiles). Research of Lahiri is partially supported by NSF grant no. DMS-0072571. Research of Mukherjee is partially supported by the Academic Research Grant R-155-000-003-112 from the National University of Singapore.     

Using DEA and simulation in guiding operating units to improved performance

The potential for the use of DEA and simulation in a mutually supporting role in guiding operating units to improved performance is presented. An analysis following a three-stage process is suggested. Stage one involves obtaining the data for the DEA analysis. This can be sourced from historical data, simulated data or a combination of the two. Stage two involves the DEA analysis that identifies benchmark operating units. In the third stage simulation can now be used in order to offer practical guidance to operating units towards improved performance. This can be achieved by the use of sensitivity analysis of the benchmark unit using a simulation model to offer direct support as to the feasibility and efficiency of any variations in operating practices to be tested. Alternatively, the simulation can be used as a mechanism to transmit the practices of the benchmark unit to weaker performing units by building a simulation model of the weaker unit to the process design of the benchmark unit. The model can then compare performance of the current and benchmark process designs. Quantifying improvement in this way provides a useful driver to any process change initiative that is required to bring the performance of weaker units up to the best in class.     

Analysis of a predator-prey model with modified Leslie-Gower and Holling-type II schemes with stochastic perturbation

In this paper, we consider a predator-prey model with modified Leslie-Gower and Holling-type II schemes with stochastic perturbation. We show there is a unique positive solution to the system with positive initial value, and mainly investigate the long time behavior of the system. Condition for the system to be extinct is given and persistent condition is established. At last, numerical simulations are carried out to support our results.