The distributed simulation of clustered processes

An efficient method for use in the discrete-event distributed simulation of large systems is presented. A conservative distributed simulation as well as the mapping of several simulated processes onto the same processor is assumed. The method consits of two algorithms: an algorithm for computing the lower bounds on times of future events, and a distributed algorithm that resolves deadlocks. The performance of the method is demonstrated by comparing it to the Chandy-Misra-Bryant simulation and by presenting some experimental results. Bojan Groelj received his B.EE. and M.EE. degrees from the University of Ljubljana, Slovenia (Yugoslavia), in 1978 and 1981, respectively. His Ph.D. degree in Computer Science is from McGill University, Montreal, Canada (1988). In 1988, he joined the Center for Advanced Computer Studies, University of Southwestern Louisiana, where he is currently an Assistant Professor. His current research interests include distributed simulation, distributed computing in general, program proving, and sorting. For the last two years, the author has been proving a distributed deadlock-detection algorithm using the UNITY approach. Carl Tropper is an Associate Professor of Computer Science at McGill University, Montreal, Canada. His major area of research at present is distributed discrete-event simulation. He has also worked in the area of performance evaluation of computer networks. Formerly, he was with BBN Communications Corporation, Cambridge, Mass., where he contributed to the design and evaluation of various widearea network protocols.This research was supported in part by the National Science Foundation under Grant CCR-8909098. An extended version of this work will appear in Progress in Simulation, vol. II, Ablex Publishing Corporation     

Digital simulation of Poisson stochastic differential equations

A simple transformation between ‘ I to ’ and ‘ ordinary ’ Poisson stochastic differential equations is obtained for the scalar, time-dependent case. This is used to specify numerical simulation procedures applicable to the respective equations. Four numerical examples are given which show the relative accuracy of the different methods.     

Doubly stochastic Poisson processes in artificial neural learning

This paper investigates neuron activation statistics in artificial neural networks employing stochastic arithmetic. It is shown that a doubly stochastic Poisson process is an appropriate model for the signals in these circuits.     

Statistical inference for doubly stochastic multichannel Poisson processes: A PCA approach

Efficient computational algorithms for making inferences about the intensity process of an observed doubly stochastic multichannel Poisson process are designed. The proposed solution is based on a numerical version of principal component analysis (PCA) of stochastic processes and hence it can be applied simply with knowledge of the first- and second-order moments of the intensity process of interest. The technique provided is valid for solving all types of estimation problems: filtering, prediction and smoothing.     

Generalization bounds for non-stationary mixing processes

This paper presents the first generalization bounds for time series prediction with a non-stationary mixing stochastic process. We prove Rademacher complexity learning bounds for both average-path generalization with non-stationary \(\beta \)-mixing processes and path-dependent generalization with non-stationary \(\phi \)-mixing processes. Our guarantees are expressed in terms of \(\beta \)- or \(\phi \)-mixing coefficients and a natural measure of discrepancy between training and target distributions. They admit as special cases previous Rademacher complexity bounds for non-i.i.d. stationary distributions, for independent but not identically distributed random variables, or for the i.i.d. case. We show that, using a new sub-sample selection technique we introduce, our bounds can be tightened under the natural assumption of asymptotically stationary stochastic processes. We also prove that fast learning rates can be achieved by extending existing local Rademacher complexity analyses to the non-i.i.d. setting. We conclude the paper by providing generalization bounds for learning with unbounded losses and non-i.i.d. data.     

Linear systems with non-stationary random inputs†

This paper concerns the response correlation matrix of a linear time variable system with stationary or non-stationary stochastic inputs modulated by linear homogeneous operators. By adapting Lanczos' ‘selected points’ method an algorithm is introduced to compute the state covariance matrix of systems with deterministic initial state. Certain computational features of response correlation matrices of systems with random initial state and random inputs are also outlined. Limited numerical studies indicate that the proposed method is efficient to treat uni-dimensional and small-order multi-dimensional systems. Mean square response comparisons with some of the previously studied systems indicate good agreement and the saving in machine time appears to be substantial.     

Semiparametric stochastic frontier models for clustered data

The mixed model approach to semiparametric regression is considered for stochastic frontier models, with focus on clustered data. Standard assumptions about the model component representing the inefficiency effect lead to a closed skew normal distribution for the response. Model parameters are estimated by a generalization of restricted maximum likelihood, and random effects are estimated by an orthodox best linear unbiased prediction procedure. The method is assessed by means of Monte Carlo studies, and illustrated by an empirical application on hospital productivity.     

On large deviations for Poisson stochastic integrals

We obtain asymptotically exact estimates for large deviations of Poisson stochastic integrals. We also find a region where such an integral can be approximated by the corresponding Gaussian random variable. In of all these results, we obtain nonasymptotic estimates for remainder terms.     

基于综合系数的随机存储系统建模与仿真

讨论了需求量、需求发生的间隔均为随机变量的(s,S)策略随机存储系统。以库存总费用和缺货率为系统性能指标,建立仿真模型,将二者加权平均给出策略的综合系数的计算公式。设计了仿真运行图,在各种费用参数给定后,利用计算机仿真技术和综合系数来选定最优存储策略。     

Gaussian process approach to spiking neurons for inhomogeneous Poisson inputs.

This article presents a new theoretical framework to consider the dynamics of a stochastic spiking neuron model with general membrane response to input spike. We assume that the input spikes obey an inhomogeneous Poisson process. The stochastic process of the membrane potential then becomes a gaussian process. When a general type of the membrane response is assumed, the stochastic process becomes a Markov-gaussian process. We present a calculation method for the membrane potential density and the firing probability density. Our new formulation is the extension of the existing formulation based on diffusion approximation. Although the single Markov assumption of the diffusion approximation simplifies the stochastic process analysis, the calculation is inaccurate when the stochastic process involves a multiple Markov property. We find that the variation of the shape of the membrane response, which has often been ignored in existing stochastic process studies, significantly affects the firing probability. Our approach can consider the reset effect, which has been difficult to deal with by analysis based on the first passage time density.     

一种基于分形和相似性查找的非平稳时间序列符号化表示法

传统的时间序列表示方法均在不同程度上采用了对数据的约简手段,从而破坏了时间序列的非线性和分形这些重要的本质特征,也就使得时间序列的相似性匹配误差加大。提出一种高精度的随机非平稳时间序列表示方法FSPA,该方法将分形理论和R/S方法应用到现有的时间序列表示方法中,既保留了时间序列的非线性和分形的重要特征,同时也实现了维度的约简。实验分别在合成数据和实际数据上进行,结果表明,该方法具有更高的精度且需要较少的存储空间。     

Modeling stochastic live load for long-span bridge based on microscopic traffic flow simulation

A general framework of modeling the stochastic live load from traffic for a long-span bridge is developed. The cellular automaton (CA) traffic flow simulation technique is adopted to develop the analytical basis of the framework. Based on the traffic flow simulation results, the live load on a long-span bridge from the stochastic traffic is studied with a focus on the static component. Parametric studies of major variables of the framework, such as the length of the connecting roadways, the speed limit, and the vehicle combination, are conducted.     

Design of optimal constrained dynamic compensators for non-stationary linear stochastic systems

This paper deals with the design of fixed-order dynamic compensators for non-stationary linear stochastic systems with noisy observations, where the observation noise need not necessarily be white. An integral quadratic performance index defined over a finite time interval is employed and this yields a matrix variational problem in the compensator parameters. It is shown how the optimal, possibly time-varying, compensator parameters rnay be determined by direct solution of this variational problem using a conjugate-gradient technique. Consideration is also given to finding suboptimal compensators that are simpler to implement. In particular, an algorithm is proposed for designing compensators having gains that are constrained to be piecewise-constant functions of time, with provision for optimally choosing the instants at which gains changes occur. An illustrative numerical example is included.     

Boundary stabilisation to non-stationary solutions for deterministic and stochastic parabolic-type equations

In this work, we design explicit, finite-dimensional boundary feedback laws for stabilisation to trajectories for parabolic-type equations. The simple form of the feedback allows to write the solution of the corresponding closed-loop equation in a mild formulation via a kernel; then, taking advantage of this, the stability is shown. As an application, null stabilisation for stochastic parabolic-type equations is deduced as well. As far as we know, the present work is the first result on boundary feedback stabilisation to trajectories and for stochastic parabolic-type equations, with stability guaranteed independent of how large the level of the noise is.     

On the identification of non-stationary linear processes

Identification algorithms for non-stationary linear processes are reviewed. In control, signal processing and many other areas of applications, to track the time varying dynamics of a system is a fundamental problem. The methods considered are the recursive least squares, the Kalman filter and the stochastic approaches. Special attention is paid to the study of how different design parameters affect these algorithms. Simulation examples are shown to demonstrate the character of the trade-off between the tracking ability and noise rejection. Also, the comparison provides considerable insight into the application of these methods to the on-line identification.     

Linear filtering for bilinear stochastic differential systems with unknown inputs

Investigates the problem of state estimation for bilinear stochastic multivariable differential systems in the presence of an additional disturbance, whose statistics are completely unknown.. A linear filter is proposed, based on a suitable decomposition of the state of the bilinear system into two components. The first one is a computable function of the observations while the second component is estimated via a suitable linear filtering algorithm. No a priori information on the disturbance is required for the filter implementation. The proposed filter is robust with respect to the unknown input, in that the covariance of the estimation error is not affected by such input. Numerical simulations show the effectiveness of the proposed filter.     

基于过程模型随机仿真的TDD模块选取建模方法研究*

为了帮助项目经理合理选择TDD实施模块,基于进程代数方法对测试驱动软件开发过程和非测试驱动软件开发过程建立过程仿真模型。通过用例度量软件模块的复杂性来获取随机变量参数对模型调参,并采用该模型得到仿真结果。提出TDD模块选取算法来分析仿真结果并得出最佳TDD实施策略,最终为项目经理提供合理的决策。     

Discrete-time controller for stochastic nonlinear polynomial systems with Poisson noises

This paper presents a solution to the optimal control problem for discrete-time stochastic nonlinear polynomial systems confused with white Poisson noises subject to a quadratic criterion. The solution is obtained in the following way: a nonlinear optimal controller is first developed for polynomial systems, considering the state vector completely available for control design. Then, based on the solution of the state estimation problem for polynomial systems with white Poisson noises, the state estimate vector is used in the control law to obtain a closed-form solution. Performance of this controller is compared to that of the controller employing the extended Kalman filter and the linear-quadratic regulator and the controller designed for polynomial systems confused with white Gaussian noises.