Differentiating Through Integer Linear Programs with Quadratic Regularization and Davis-Yin Splitting

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: McKenzie, Daniel, Fung, Samy Wu, Heaton, Howard
Format: Preprint
Veröffentlicht: 2023
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866929428052836352
author McKenzie, Daniel
Fung, Samy Wu
Heaton, Howard
author_facet McKenzie, Daniel
Fung, Samy Wu
Heaton, Howard
contents In many applications, a combinatorial problem must be repeatedly solved with similar, but distinct parameters. Yet, the parameters $w$ are not directly observed; only contextual data $d$ that correlates with $w$ is available. It is tempting to use a neural network to predict $w$ given $d$. However, training such a model requires reconciling the discrete nature of combinatorial optimization with the gradient-based frameworks used to train neural networks. We study the case where the problem in question is an Integer Linear Program (ILP). We propose applying a three-operator splitting technique, also known as Davis-Yin splitting (DYS), to the quadratically regularized continuous relaxation of the ILP. We prove that the resulting scheme is compatible with the recently introduced Jacobian-free backpropagation (JFB). Our experiments on two representative ILPs: the shortest path problem and the knapsack problem, demonstrate that this combination-DYS on the forward pass, JFB on the backward pass-yields a scheme which scales more effectively to high-dimensional problems than existing schemes. All code associated with this paper is available at github.com/mines-opt-ml/fpo-dys.
format Preprint
id arxiv_https___arxiv_org_abs_2301_13395
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Differentiating Through Integer Linear Programs with Quadratic Regularization and Davis-Yin Splitting
McKenzie, Daniel
Fung, Samy Wu
Heaton, Howard
Machine Learning
In many applications, a combinatorial problem must be repeatedly solved with similar, but distinct parameters. Yet, the parameters $w$ are not directly observed; only contextual data $d$ that correlates with $w$ is available. It is tempting to use a neural network to predict $w$ given $d$. However, training such a model requires reconciling the discrete nature of combinatorial optimization with the gradient-based frameworks used to train neural networks. We study the case where the problem in question is an Integer Linear Program (ILP). We propose applying a three-operator splitting technique, also known as Davis-Yin splitting (DYS), to the quadratically regularized continuous relaxation of the ILP. We prove that the resulting scheme is compatible with the recently introduced Jacobian-free backpropagation (JFB). Our experiments on two representative ILPs: the shortest path problem and the knapsack problem, demonstrate that this combination-DYS on the forward pass, JFB on the backward pass-yields a scheme which scales more effectively to high-dimensional problems than existing schemes. All code associated with this paper is available at github.com/mines-opt-ml/fpo-dys.
title Differentiating Through Integer Linear Programs with Quadratic Regularization and Davis-Yin Splitting
topic Machine Learning
url https://arxiv.org/abs/2301.13395