Saved in:
Bibliographic Details
Main Authors: Enaieh, Ikhlas, Fercoq, Olivier
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2603.04133
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908866057338880
author Enaieh, Ikhlas
Fercoq, Olivier
author_facet Enaieh, Ikhlas
Fercoq, Olivier
contents Deep Neural Networks are powerful tools for solving machine learning problems, but their training often involves dense and costly parameter updates. In this work, we use a novel Max-Plus neural architecture in which classical addition and multiplication are replaced with maximum and summation operations respectively. This is a promising architecture in terms of interpretability, but its training is challenging. A particular feature is that this algebraic structure naturally induces sparsity in the subgradients, as only neurons that contribute to the maximum affect the loss. However, standard backpropagation fails to exploit this sparsity, leading to unnecessary computations. In this work, we focus on the minimization of the worst sample loss which transfers this sparsity to the optimization loss. To address this, we propose a sparse subgradient algorithm that explicitly exploits the algebraic sparsity. By tailoring the optimization procedure to the non-smooth nature of Max-Plus models, our method achieves more efficient updates while retaining theoretical guarantees. This highlights a principled path toward bridging algebraic structure and scalable learning.
format Preprint
id arxiv_https___arxiv_org_abs_2603_04133
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Exploiting Subgradient Sparsity in Max-Plus Neural Networks
Enaieh, Ikhlas
Fercoq, Olivier
Machine Learning
Deep Neural Networks are powerful tools for solving machine learning problems, but their training often involves dense and costly parameter updates. In this work, we use a novel Max-Plus neural architecture in which classical addition and multiplication are replaced with maximum and summation operations respectively. This is a promising architecture in terms of interpretability, but its training is challenging. A particular feature is that this algebraic structure naturally induces sparsity in the subgradients, as only neurons that contribute to the maximum affect the loss. However, standard backpropagation fails to exploit this sparsity, leading to unnecessary computations. In this work, we focus on the minimization of the worst sample loss which transfers this sparsity to the optimization loss. To address this, we propose a sparse subgradient algorithm that explicitly exploits the algebraic sparsity. By tailoring the optimization procedure to the non-smooth nature of Max-Plus models, our method achieves more efficient updates while retaining theoretical guarantees. This highlights a principled path toward bridging algebraic structure and scalable learning.
title Exploiting Subgradient Sparsity in Max-Plus Neural Networks
topic Machine Learning
url https://arxiv.org/abs/2603.04133