A regularized cutting plane algorithm for inverse mixed integer programming problem

This repository contains the code and instances associated with the following paper

"A regularized cutting plane algorithm for inverse mixed integer programming Problem"

by Vishnu Vijayaraghavan, Kiavash Kianfar, Andrew Schaefer, submitted, 2019

Implemented Algorithms

1. Cutting Plane Algorithm presented in "Wang, L., Cutting plane algorithms for the inverse mixed integer linear programming problem, Operations Research Letters 37: 114-116, 2009

2. Regularized Cutting Plane Algorithm developed in "Vijayaraghavan, V., Kianfar, K., Schaefer, A., A regularized cutting plane algorithm for inverse mixed integer programming Problem" submitted, 2019

Code

Implementation in C++ with Optimization Problems solved using ILOG CPLEX 12.9 callable library on Windows Platform

Files :

1. ProbMIP.cpp :

   Sets up the MIP/ Regularized MIP problems

   Implements functions to randomly generate Inverse MIP instances based on the Seed obtained

   Includes functions to solve the MIP instances and obtain the solutions

2. GenLP.cpp:

   Sets up the Cost Generator LP problem

   Includes functions to add new cuts to the LP problem instance

   Also includes functions to solve the LP instances and obtain the solutions

3. Main.cpp:

   Includes the main() function that implements the algorithms

   Takes the problem name as a command line argument while execution

The variable seed needs to be set to a positive integer value for the algorithm to generate random instances of the Inverse MIP problem

Instructions to Execute the Code

1. Include the files Main.cpp, ProbMIP.cpp, ProbMIP.h, GenLP.cpp, and GenLp.h

2. Pass the MIP problem to be used as a command line argument when calling the executable

3. The variable seed in Main.cpp to be set to a positive integer to generate a random instance, a random instances being a combination of randomly generated objective coefficient and required solution for the Inverse MIP problem

4. To execute the Cutting Plane Algorithm include the directive #define SOLVE_MIPCUT before the main() function in Main.cpp

5. To execute the Regularized Cutting Plane Algorithm include the directive #define SOLVE_MIPREG before the main() function in Main.cpp

Instances

Instances.zip contains MIP instances form MIPLIB 2003 and MIPLIB2010 which are used in the paper to generate random inverse MIP instances.