On interior-point based retrospective approximation methods for solving two-stage stochastic linear programs

Authors:
Soumyadip Ghosh;Raghu Pasupathy
Affiliations:
IBM T.J. Watson Research Center, Yorktown Heights, NY;Virginia Tech, Blacksburg, VA
Venue:
Proceedings of the Winter Simulation Conference
Year:
2011

Citing 9
Cited 1

Asymptotic theory for solutions in statistical estimation and stochastic programming

Mathematics of Operations Research
Retrospective approximation algorithms for stochastic root finding

WSC '94 Proceedings of the 26th conference on Winter simulation
Single-Period Multiproduct Inventory Models with Substitution

Operations Research
Convex Optimization

Convex Optimization
On Rates of Convergence for Stochastic Optimization Problems Under Non-Independent and Identically Distributed Sampling

SIAM Journal on Optimization
Retrospective-approximation algorithms for the multidimensional stochastic root-finding problem

ACM Transactions on Modeling and Computer Simulation (TOMACS)
Optimal Allocation of Surgery Blocks to Operating Rooms Under Uncertainty

Operations Research
On Choosing Parameters in Retrospective-Approximation Algorithms for Stochastic Root Finding and Simulation Optimization

Operations Research
A Sequential Sampling Procedure for Stochastic Programming

Operations Research

An introspective on the retrospective-approximation paradigm

Proceedings of the Winter Simulation Conference

Quantified Score

Hi-index	0.00

Visualization

Abstract

In a recent paper, Gongyun Zhao introduced what appears to be the first interior point formulation for solving two-stage stochastic linear programs for finite support random variables. In this paper, we generalize Gongyun Zhao's formulation by incorporating it into a retrospective approximation framework. What results is an implementable interior-point solution paradigm that can be used to solve general two-stage stochastic linear programs. After discussing some basic properties, we characterize the complexity of the algorithm, leading to guidance on the number of samples that should be generated to construct the sub-problem linear programs, effort expended in solving the sub-problems, and the effort expended in solving the master problem.