Skip to main navigation Skip to search Skip to main content

ProdNet: A Lightweight Network for Fast Discovery of Matrix Multiplication Algorithms

  • Badia Ouissam Lakas*
  • , Chemousse Berdjouh
  • , Khadra Bounane
  • , Mohammed Lamine Kherfi
  • , Oussama Aiadi
  • , Samir Brahim Belhouari
  • *Corresponding author for this work
  • University Kasdi Merbah Ouargla
  • Sultan Qaboos University

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Matrix multiplication is a fundamental operation with significant applications across diverse fields, such as physics, electronics, and artificial intelligence. Traditional implementations of this operation exhibit cubic time complexity, which presents computational challenges, particularly in deep learning scenarios that necessitate large-scale matrix computations. In this study, we introduce ProdNet, a lightweight neural network model designed to autonomously discover efficient matrix multiplication algorithms without the need for extensive computational resources or prior knowledge of existing methods. Our approach seeks to alleviate complexity by minimizing the number of multiplicative operations involved. To achieve this, we utilize a combination of Mean Squared Error (MSE) and a regularization function, targeting weight values to be constrained to 0, 1, or –1. We emphasize the effectiveness of our regularization techniques in accelerating the algorithm discovery process.

Original languageEnglish
Title of host publicationIntelligent Systems and Pattern Recognition - 5th International Conference, ISPR 2025, Revised Selected Papers
EditorsTolga Ensari, Akram Bennour, Bassem Bouaziz, Walid Mahdi, Imad Rida
PublisherSpringer Science and Business Media Deutschland GmbH
Pages95-108
Number of pages14
ISBN (Print)9783032215840
DOIs
Publication statusPublished - 2026
Event5th International Conference on Intelligent Systems and Pattern Recognition, ISPR 2025 - Hammamet, Tunisia
Duration: 25 Sept 202527 Sept 2025

Publication series

NameCommunications in Computer and Information Science
Volume2859 CCIS
ISSN (Print)1865-0929
ISSN (Electronic)1865-0937

Conference

Conference5th International Conference on Intelligent Systems and Pattern Recognition, ISPR 2025
Country/TerritoryTunisia
CityHammamet
Period25/09/2527/09/25

Keywords

  • Matrix product algorithm
  • neural network
  • ProdNet
  • regularization

Fingerprint

Dive into the research topics of 'ProdNet: A Lightweight Network for Fast Discovery of Matrix Multiplication Algorithms'. Together they form a unique fingerprint.

Cite this