Skip to content

The repository hosts the codes developed to implement decision rules that are used to solve multi-stage robust optimization problems

Notifications You must be signed in to change notification settings

guxu-iowa/OR-MSRO

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

17 Commits
 
 
 
 
 
 
 
 

Repository files navigation

OR-MSRO

Background

The repository hosts the codes developed to implement decision rules that are used to solve multi-stage robust optimization problems. The decision rules are developed in the paper "Improved Decision Rule Approximations for Multi-Stage Robust Optimization via Copositive Programming".

All optimization problems are solved using MOSEK 8.1.0.56 via the YALMIP interface on a 16-core 3.4 GHz Linux PC with 32 GB RAM. Computation times are reported in the paper.

The repository has three folders with one for each of the three examples in the paper.

Newsvendor

File test.m is the entrance of the program. It generates test instances and run different decision rules over those instances. Decision rules include quadratic decision rule (newsvendor_qdr.m), copositive-based decision rule (newsvendor_cop.m), approximate s_lemma (newsvendor_as.m), and polynomial decision rule (newsvendor_poly.m). Auxiliary files including genBC.m, genind.m,and genpow.m are used to implement polynomial decision rule.

Inventory_control

File test.m is the entrance of the program. The file calls the file "run_decision_rules.m", which actually generates inventory control instances and compare different decision rules over these instances. The compared decision rules are implemented in approxSL.m (aproximate S-Lemma decision rule), ldr.m (linear decision rule), piecewiseLDR.m (piecewise linear decision rule) and pdr.m (polynomial decision rule) with degrees 2 and 3.

Index_tracking

File test.m is the entrance of the program. The file calls the file "run_decision_rules.m", which actually generates index tracking instances and compare different decision rules over these instances. The compared decision rules are implemented in approxSL.m (aproximate S-Lemma decision rule), lqdr.m (linear and quadratic decision rule), and pdr.m (polynomial decision rule) with degrees 2 and 3.

References

ApS M (2016) The MOSEK optimization toolbox for MATLAB manual. Version 8.0. URL http://docs. mosek.com/8.0/toolbox/index.html

Lofberg J (2004) Yalmip: A toolbox for modeling and optimization in matlab. Computer Aided Control Systems Design, 2004 IEEE International Symposium on, 284–289 (IEEE)

About

The repository hosts the codes developed to implement decision rules that are used to solve multi-stage robust optimization problems

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages