A computational status update for exact rational mixed integer programming

The last milestone achievement for the roundoff-error-free solution of general mixed integer programs over the rational numbers was a hybrid-precision branch-and-bound algorithm published by Cook, Koch, Steffy, and Wolter in 2013. We describe a substantial revision and extension of this framework that integrates symbolic presolving, features an exact repair step for solutions from primal heuristics, employs a faster rational LP solver based on LP iterative refinement, and is able to produce independently verifiable certificates of optimality. We study the significantly improved performance and give insights into the computational behavior of the new algorithmic components. On the MIPLIB 2017 benchmark set, we observe an average speedup of 10.7x over the original framework and 2.9 times as many instances solved within a time limit of two hours.

Automated summary

A new algorithmic framework for general mixed integer programming problems is proposed, which integrates symbolic presolving, exact repair steps, and a faster LP solver. This framework shows significantly improved performance on benchmark instances, with an average speedup of 10.7x and 2.9 times as many instances solved within a time limit of two hours compared to the original framework.