The two-moment three-parameter decomposition approximation of queueing networks with exponential residual renewal processes

We propose a two-moment three-parameter decomposition approximation of general open queueing networks by which both autocorrelation and cross correlation are accounted for. Each arrival process is approximated as an exponential residual (ER) renewal process that is characterized by three parameters: intensity, residue, and decrement. While the ER renewal process is adopted for modeling autocorrelated processes, the innovations method is used for modeling the cross correlation between randomly split streams. As the interarrival times of an ER renewal process follow a two-stage mixed generalized Erlang distribution, viz., MGE(2), each station is analyzed as an MGE(2)/G/1 system for the approximate mean waiting time. Variability functions are also used in network equations for a more accurate modeling of the propagation of cross correlations in queueing networks. Since an ER renewal process is a special case of a Markovian arrival process (MAP), the value of the variability function is determined by a MAP/MAP/1 approximation of the departure process. Numerical results show that our proposed approach greatly improves the performance of the parametric decomposition approximation of open queueing networks.     

Efficient and robust estimation for autoregressive regression models using shape mixtures of skewt normal distribution

Multiple linear regression model based on normally distributed and uncorrelated errors is a popular statistical tool with application in various fields. But these assumptions of normality and no serial correlation are hardly met in real life. Hence, this study considers the linear regression time series model for series with outliers and autocorrelated errors. These autocorrelated errors are represented by a covariance-stationary autoregressive process where the independent innovations are driven by shape mixture of skew-t normal distribution. The shape mixture of skew-t normal distribution is a flexible extension of the skew-t normal with an additional shape parameter that controls skewness and kurtosis. With this error model, stochastic modeling of multiple outliers is possible with an adaptive robust maximum likelihood estimation of all the parameters. An Expectation Conditional Maximization Either algorithm is developed to carryout the maximum likelihood estimation. We derive asymptotic standard errors of the estimators through an information-based approximation. The performance of the estimation procedure developed is evaluated through Monte Carlo simulations and real life data analysis.

     

多元自相关过程的VAR控制图

为了解决多元自相关过程的残差T~2控制图对小偏移不灵敏的问题,本文利用批量-均值法的思想,结合VAR模型的渐近分布,设计了多元自相关过程的向量自回归(VAR)控制图.只要子组样本量足够大,VAR控制图可以对过程出现的各种偏移进行有效控制.通过对比残差T~2控制图的控制效果,得出VAR控制图对小偏移灵敏、残差T~2控制图对大偏移灵敏的结论,联合使用VAR控制图和残差T~2控制图可更有效地监控多元自相关过程。     

自相关对常规控制图影响的模拟研究与案例分析

常规统计控制图的基本假设前提是观测值独立同分布,而在实际生产过程中,质量指标值常表现出自相关现象,违背独立性假定。本文运用平均链长(ARL)研究自相关过程为AR(1)时对常规控制图的影响,并比较了常规控制图和残差控制图对序列相关过程的控制效果。模拟结果和实例分析表明:当过程序列相关时,使用常规作图法估计出的标准差是有偏的,致使控制限设置错误和常规控制图检测能力降低。因此,在一些统计过程控制中,须考虑自相关现象并采用适当的控制图方法。     

自相关过程的残差控制图

由于自相关过程违背了过程输出数据独立性的假定,使得传统休哈特图的有效性受到质疑。本文首先讨论控制图设计基本思想,然后分析了对自相关过程监控的残差控制图原理;进而以平均链长和各链点检出概率为准则,系统研究了AR(1)过程残差控制图的检测能力,并与休图进行了比较。最后,通过一个模拟验证了该方法的有效性。     

On the estimation of optimal batch sizes in the analysis of simulation output

The estimation of the variance of point estimators is a classical problem of stochastic simulation. A more specific problem addresses the estimation of the variance of a sample mean from a steady-state autocorrelated process. Many proposed estimators of the variance of the sample mean are parameterized by batch size. A critical problem is to find an appropriate batch size that provides a good tradeoff between bias and variance. This paper proposes a procedure for determining the optimal batch size to minimize the mean squared error of estimators of the variance of the sample mean. This paper also presents the results of empirical studies of the procedure. The experiments involve symmetric two-state Markov chain models, first-order autoregressive processes, seasonal autoregressive processes, and queue-waiting times for several M/M/1 queueing models. The empirical results indicate that the estimation procedure works nearly as well as it would if the parameters of the processes were known.     

Queues with Dropping Functions and Autocorrelated Arrivals

We present an analysis of the queueing system in which arriving jobs are dropped with probability depending on the queue size. The arrivals are assumed to be autocorrelated and they are modeled by the Markov-modulated Poisson process. Both transient and stationary distributions of the queue size, as well as the system loss ratio and throughput are obtained. The analytical results are accompanied with numerical examples based on the autocorrelated traffic recorded in an IP computer network.     

Generating pseudo-random time series with specified marginal distributions

Our goal is to generate a target time series with a specified marginal distribution and a specified lag-one autocorrelation. We consider an existing approach: first transform a known autocorrelated reference series into the corresponding uniform autocorrelated series and then apply the specified inverse transformation to each observation producing the target series. This approach is simple, except that the lag-one reference-series autocorrelation must be determined in a set-up step. We propose a method for determining this autocorrelation.     

Approximation processes for fourier-jacobi expansions †

This paper deals with the approximation theoretic aspects of summation methods for expansions in terms of Jacobi polynomials. When a funcation f is expanded in a Fourier-Jacobi series, many summation methods for this series may be looked upon as approximation processes for the function f. The main object of this paper is to investigate the order of approximation of these processes and to characterize the functions which allow a certain order of approximation. Many of these processes exhibit the phenomenon of saturation, which is equivalent to the existence of an optimal order of approximation (the saturation, which is equivalent to the existence of an optimal order of approximation (the saturation order). For the approximation processes treated in this paper the saturation order and the saturation class, that is the class if functions which can be approximated with the optimal order, are derived. The characterization of the classes of functions is accomplished by means of the theory of intermediate spaces due to Peetre[19] (compare Butzer and Berens [7]). Another basic tool in this work is the convolution structure for Jacobi series, introduced by Askey and Wainger [1] (see also Gasper [14], {15})     

Binary Recurrent Statistical Experiments with Persistent Linear Regression

We construct a stochastic approximation of binary statistical experiments with persistent regression by the autoregression process with normal perturbations. The continuous-time diffusion approximation of the Ornstein–Uhlenbeck type is obtained. The stationary distributions of the approximation processes are calculated.     

Strong converse inequalities

Techniques are developed to obtain strong converse inequalities for various linear approximation processes. This will establish equivalence between the approximating rate of a certain linear process and the appropriate PeetreK-functional. Approximation processes that will be treated have to be saturated asK-functionals are saturated. These general methods will lead to new results on the various trigonometric polynomial approximation processes, on holomorphic semigroups, on Bernstein polynomials and on other commonly used approximation processes. Supported by NSERC A4816 of Canada.     

Mean Waiting Time Approximations in the G/G/1 Queue

It is known that correlations in an arrival stream offered to a single-server queue profoundly affect mean waiting times as compared to a corresponding renewal stream offered to the same server. Nonetheless, this paper uses appropriately constructed GI/G/1 models to create viable approximations for queues with correlated arrivals. The constructed renewal arrival process, called PMRS (Peakedness Matched Renewal Stream), preserves the peakedness of the original stream and its arrival rate; furthermore, the squared coefficient of variation of the constructed PMRS equals the index of dispersion of the original stream. Accordingly, the GI/G/1 approximation is termed PMRQ (Peakedness Matched Renewal Queue). To test the efficacy of the PMRQ approximation, we employed a simple variant of the TES⁺ process as the autocorrelated arrival stream, and simulated the corresponding TES ⁺/G/1 queue for several service distributions and traffic intensities. Extensive experimentation showed that the proposed PMRQ approximations produced mean waiting times that compared favorably with simulation results of the original systems. Markov-modulated Poisson process (MMPP) is also discussed as a special case.     

Simulation of stochastic integrals with respect to Lévy processes of type G

We study the simulation of stochastic processes defined as stochastic integrals with respect to type G Lévy processes for the case where it is not possible to simulate the type G process exactly. The type G Lévy process as well as the stochastic integral can on compact intervals be represented as an infinite series. In a practical simulation we must truncate this representation. We examine the approximation of the remaining terms with a simpler process to get an approximation of the stochastic integral. We also show that a stochastic time change representation can be used to obtain an approximation of stochastic integrals with respect to type G Lévy processes provided that the integrator and the integrand are independent.     

On the asymptotics of locally dependent point processes

We investigate a family of approximating processes that can capture the asymptotic behaviour of locally dependent point processes. We prove two theorems presented to accommodate respectively the positively and negatively related dependent structures. Three examples are given to illustrate that our approximating processes can circumvent the technical difficulties encountered in compound Poisson process approximation (see Barbour and Månsson (2002) [10]) and our approximation error bound decreases when the mean number of the random events increases, in contrast to the increasing of bounds for compound Poisson process approximation.     

Approximation for bootstrapped empirical processes

We obtain an approximation for the bootstrapped empirical process with the rate of the Komlós, Major and Tusnády approximation for empirical processes. The proof of the new approximation is based on the Poisson approximation for the uniform empirical distribution function and the Gaussian approximation for randomly stopped sums.

     

一阶自回归模型参数变点的假设检验

本文讨论一阶自回归模型自回归参数$\phi$的变点问题. 对于一阶自回归模型, 在模型的白噪声序列的方差$\sigma^2$已知和未知的条件下, 利用最大似然方法, 我们分别讨论了模型自回归参数$\phi$的Abrupt Change-Point 和Gradual Change-Point的检测问题.     

Martingale approximations for continuous-time and discrete-time stationary Markov processes

We show that the method of Kipnis and Varadhan [Comm. Math. Phys. 104 (1986) 1–19] to construct a Martingale approximation to an additive functional of a stationary ergodic Markov process via the resolvent is universal in the sense that a martingale approximation exists if and only if the resolvent representation converges. A sufficient condition for the existence of a martingale approximation is also given. As examples we discuss moving average processes and processes with normal generator.     

The evolutionary dynamics of a spatial multi-strain host-pathogen system with cross-immunity

We considered a Susceptible-Infective-Recovered-Susceptible (SIRS) model with strain mutation and cross-immunity in a non-spatial model and a lattice-structured model, where all individuals can reproduce if the space/resources allow. In the lattice-structured model, both the host reproduction and pathogen transmission processes are assumed to interact with next nearest neighbors, and the model was analyzed by an improved pair approximation (IPA). A family of correlated equations of pair approximation and mean-field were presented. We show the phase diagram of the coexistence and extinction which were obtained from parameterization by measuring the basic reproduction numbers of the strains during their infection processes. The qualitative results of the pair approximation model are similar to that of the corresponding non-spatial model. Furthermore, the spatial model predicts coexistence over a wider range of parameters than the non-spatial model. In particular, when the strain evolution tends to a larger basic reproduction number, the correlated spatial approximation could predict better than the mean-field approximation.     

On the rate of convergence of simple and jump-adapted weak Euler schemes for Lévy driven SDEs

The paper studies the rate of convergence of a weak Euler approximation for solutions to possibly completely degenerate SDEs driven by Lévy processes, with Hölder-continuous coefficients. It investigates the dependence of the rate on the regularity of coefficients and driving processes and its robustness to the approximation of the increments of the driving process. A convergence rate is derived for some approximate jump-adapted Euler scheme as well.     

Strong approximations of semimartingales by processes with independent increments

Summary Strong approximation theorems for continuous time semimartingales are obtained by combining some techniques of the general theory of stochastic processes with some of the direct approximation of dependent random variables by independent ones. Continuous processes with independent increments whose variance functions increase polynomially or exponentially are considered as approximating processes. The basic assumptions of the main results only contain rates of convergence for certain probabilities. In particular, moment assumptions are not required. Some almost sure invariance principles for partial sum processes with nonlinear growth of variance and for functionals of Markov processes are derived by applying the main results.