STOCHASTIC LINEAR PROGRAMMING WITH RECOURSE: A TUTORIAL*

Presented here is an introduction to stochastic linear programs with recourse. The paper is by no means a comprehensive survey of the field; that would be an encyclopedic task at best. This paper discusses formulation, interpretation, and computational aspects of stochastic linear programs with simple and fixed recourse. The paper is pedagogical in nature and is aimed to whet the interest of the decision scientist that has little or no background in stochastic programming. Moreover, the papers in this field have appeared in diverse journals not always readily available to the typical management scientist or practitioner and quite often at a very sophisticated mathematical level. For the practitioner, Wets's algorithm for solving stochastic linear programs with simple recourse may be particularly interesting since Wets shows in that paper how the problem can be reduced to an equivalent deterministic linear program of the same dimensionality.     

Noisy Stochastic Games

This paper establishes existence of a stationary Markov perfect equilibrium in general stochastic games with noise—a component of the state that is nonatomically distributed and not directly affected by the previous period's state and actions. Noise may be simply a payoff‐irrelevant public randomization device, delivering known results on the existence of correlated equilibrium as a special case. More generally, noise can take the form of shocks that enter into players' stage payoffs and the transition probability on states. The existence result is applied to a model of industry dynamics and to a model of dynamic electoral competition.     

Windows环境下实现实时数据采集的可行性

分析了Ｗｉｎｄｏｗｓ系统的消息驱动及多任务机制的实质，指出了Ｗｉｎｄｏｗｓ系统在实时性方面的所存在的不足，结合动态链接库的特点，给出了一种在Ｗｉｎｄｏｗｓ环境下实现实时数据采集的途径，即软件开发使用动态链接库。     

A Purchasing Framework for B2B Pricing Decisions and Risk‐sharing in Supply Chains*

This paper presents a common modelling structure for (i) the implementation of operational policies by individual purchasing managers of risk‐sharing agreements among supply‐chain partners, and (ii) the integration of brick and click purchasing policies in a B2B. The problem of price uncertainty created within these two environments is modelled as a stochastic repetitive‐sales problem, applicable to any probability distribution. The model identifies sufficient conditions for regenerative ordering cycles, which allows for the use of the renewal reward theorem. The end result is a two‐price purchasing policy, which may substantially ease implementation problems across a global corporation's purchasing managers world‐wide and across B2B markets.