Split Cuts for Stochastic Integer Programming

With stochastic integer programming as the motivating application, we investigate techniques to use integrality information to obtain improved cuts within a Benders decomposition algorithm. We consider two options: (i) cut-and-project, where integrality information is used to derive cuts in the extended variable space, and Benders cuts are then used to project the resulting improved relaxation, and (ii) project-and-cut, where integrality information is used to derive cuts directly in the projected space defined by Benders cuts. We analyze the use of split cuts in these two approaches, and demonstrate that although they yield equivalent relaxations when considering a single split, cut-and-project yields stronger relaxations in general when using multiple splits. Computational results illustrate that the difference can be very large, and demonstrate that using split cuts within the cut-and-project framework can significantly outperform other general purpose methods.

By: Merve Bodur, Sanjeeb Dash, Oktay Günlük, James Luedtke

Published in: RC25427 in 2013


This Research Report is available. This report has been submitted for publication outside of IBM and will probably be copyrighted if accepted for publication. It has been issued as a Research Report for early dissemination of its contents. In view of the transfer of copyright to the outside publisher, its distribution outside of IBM prior to publication should be limited to peer communications and specific requests. After outside publication, requests should be filled only by reprints or legally obtained copies of the article (e.g., payment of royalties). I have read and understand this notice and am a member of the scientific community outside or inside of IBM seeking a single copy only.


Questions about this service can be mailed to reports@us.ibm.com .