StAMPL: A Filtration-Oriented Modeling Tool for Multistage Stochastic Recourse Problems

Authors:
Robert Fourer;Leo Lopes
Affiliations:
Department of Industrial Engineering and Management Sciences, Northwestern University, Evanston, Illinois 60208;Systems and Industrial Engineering Department, University of Arizona, Tucson, Arizona 85721
Venue:
INFORMS Journal on Computing
Year:
2009

Citing 10
Cited 1

Literate programming

The Computer Journal
A modeling language for mathematical programming

Management Science
MSLiP: a computer code for the multistage stochastic linear programming problem

Mathematical Programming: Series A and B
Scenarios and policy aggregation in optimization under uncertainty

Mathematics of Operations Research
Stochastic decomposition: an algorithm for two-state linear programs with recourse

Mathematics of Operations Research
SLP-IOR: an interactive model management system for stochastic linear programs

Mathematical Programming: Series A and B
Financial Asset-Pricing Theory and Stochastic Programming Models for Asset/ Liability Management: a Synthesis

Management Science
The NEOS Server

IEEE Computational Science & Engineering
On a New Collection of Stochastic Linear Programming Test Problems

INFORMS Journal on Computing
AIMMS - Optimization Modeling

AIMMS - Optimization Modeling

Robust Optimization Made Easy with ROME

Operations Research

Quantified Score

Hi-index	0.00

Visualization

Abstract

Every multistage stochastic programming problem with recourse (MSPR) contains a filtration process. In this research, we have created a notation that makes the filtration process the central syntactic construction of the MSPR. As a result, we achieve lower redundancy and higher modularity than is possible using the mathematical notation commonly associated with stochastic programming. To experiment with our ideas, we have created StAMPL, a specialized modeling tool for the MSPR that is based on the AMPL modeling language. StAMPL implements our notation and converts models written in our notation to instances that can be solved using standard methods. Using this approach, we are able to represent models in a very clean, simple, and scalable format while maintaining almost all the power of the AMPL language.