Bounds for Two-Stage Stochastic Programs with Fixed Recourse

Abstract

This paper develops upper and lower bounds on two-stage stochastic linear programs using limited moment information. The case considered is when both the right-hand side as well as the objective coefficients of the second stage problem are random. Random variables are allowed to have arbitrary multivariate probability distributions with bounded support. First, upper and lower bounds are obtained using first and cross moments, from which we develop bounds using only first moments. The bounds are shown to solve the respective general moment problems.

Multiperiod Portfolio Optimization with Terminal Liability: Bounds for the Convex Case

“…Other methods for solving large stochastic optimization problems are generally addressed within sampling based techniques or bound based techniques. See, for instance, Ermoliev and Wets [15], Ermoliev and Gairvoronsky [14], Higle and Sen [20], Dantzig and Glynn [6] for the former technique; Birge and Wets [3,4], Edirisinghe and Ziemba [11][12][13], Frauendorfer [16,17] for the latter technique. Noting that the problem to be solved in each stage involves a constraint recourse matrix that is random, many of the standard bounding techniques within stochastic programming are inapplicable in this case.…”

Multistage stochastic programming: Error analysis for the convex case

“…This is done on the basis of barycentric approximation [19] and amounts to approximate the marginal and regular conditional probability measures which are induced through P. Barycentric approximation resembles the classical inequalities due to Jensen and Edmundson-Madansky: these inequalities have proved extremely useful in designing solution methods for stochastic programs with recourse (see e.g. [4,5,13,14,16,18,25,28,29,30]). …”

An Inexact Two-Stage Stochastic Programming Model for Sustainable Utilization of Water Resource in Dalian City

“…Wang and Adams (1986) proposed a framework of two-stage programming in response to the optimization of reservoir operation [10]. Edirisinghe and Ziemba (1994) proposed lower and upper bounds within two-stage stochastic linear programming (TSLP) through limited moment information [11]. Barik et al (2014) developed a two-stage stochastic linear programming model considering some parameters as multi-choice parameters and others as exponential random variables [12].…”