Faster integer-feasibility in mixed-integer linear programs by branching to force change

Authors:
Jennifer Pryor;John W. Chinneck
Affiliations:
Systems and Computer Engineering, Carleton University, Ottawa, Ontario, Canada K1S 5B6;Systems and Computer Engineering, Carleton University, Ottawa, Ontario, Canada K1S 5B6
Venue:
Computers and Operations Research
Year:
2011

Citing 8
Cited 4

MIP Models and BB Strategies in Brachytherapy Treatment Optimization

Journal of Global Optimization
Progress in Linear Programming-Based Algorithms for Integer Programming: An Exposition

INFORMS Journal on Computing
Octane: A New Heuristic for Pure 0-1 Programs

Operations Research
The feasibility pump

Mathematical Programming: Series A and B
Handbook of Mathematical Functions, With Formulas, Graphs, and Mathematical Tables,

Handbook of Mathematical Functions, With Formulas, Graphs, and Mathematical Tables,
Active-constraint variable ordering for faster feasibility of mixed integer linear programs

Mathematical Programming: Series A and B
Counting solutions of knapsack constraints

CPAIOR'08 Proceedings of the 5th international conference on Integration of AI and OR techniques in constraint programming for combinatorial optimization problems
Branching rules revisited

Operations Research Letters

On counting lattice points and Chvátal-Gomory cutting planes

CPAIOR'11 Proceedings of the 8th international conference on Integration of AI and OR techniques in constraint programming for combinatorial optimization problems
Three Ideas for the Quadratic Assignment Problem

Operations Research
Achieving MILP feasibility quickly using general disjunctions

Computers and Operations Research
Backdoor Branching

INFORMS Journal on Computing

Quantified Score

Hi-index	0.01

Visualization

Abstract

Branching in mixed-integer (or integer) linear programming requires choosing both the branching variable and the branching direction. This paper develops a number of new methods for making those two decisions either independently or together with the goal of reaching the first integer-feasible solution quickly. These new methods are based on estimating the probability of satisfying a constraint at the child node given a variable/direction pair. The surprising result is that the first integer-feasible solution is usually found much more quickly when the variable/direction pair with the smallest probability of satisfying the constraint is chosen. This is because this selection forces change in many candidate variables simultaneously, leading to an integer-feasible solution sooner. Extensive empirical results are given.