ScholarMate
English简体中文繁體中文
客服热线:400-1616-289
登录注册
期刊论文

The decimation scheme for symmetric matrix factorization

Camilli, Francesco*; Mezard, Marc
Journal of Physics A: Mathematical and Theoretical2024-2-23DOI: 10.1088/1751-8121/ad2299
1被引1阅读
引用收藏
分享

摘要

Matrix factorization is an inference problem that has acquired importance due to its vast range of applications that go from dictionary learning to recommendation systems and machine learning with deep networks. The study of its fundamental statistical limits represents a true challenge, and despite a decade-long history of efforts in the community, there is still no closed formula able to describe its optimal performances in the case where the rank of the matrix scales linearly with its size. In the present paper, we study this extensive rank problem, extending the alternative 'decimation' procedure that was recently introduced by Camilli and Mezard, and carry out a thorough study of its performance. Decimation aims at recovering one column/line of the factors at a time, by mapping the problem into a sequence of neural network models of associative memory at a tunable temperature. The main advantage of decimation is that its theoretical performance can be studied using statistical physics techniques. In this paper we provide the general analysis applying to a large class of compactly supported priors and we show that the replica symmetric free entropy of the neural network models takes a universal form in the low temperature limit. We then study the decimation phase diagrams in two concrete cases, the sparse Ising prior and the uniform prior, for which we show that matrix factorization is theoretically possible when the ration P/N of rank to variable is below a certain threshold. This suggest that a possible route to solving algorithmically the matrix factorization problem could be to find efficient decimation algorithms; in this respect, we show that simple simulated annealing, though being effective for limited signal sizes, is not scalable.

学科与关键词

关键词
matrix factorizationHopfield modeldecimation schemehigh-dimensional inferenceassociative memories
相关
加载中…
  • A survey on deep matrix factorizations
    De Handschutter, Pierre*; Gillis, Nicolas; Siebert, XavierComputer Science Review, 2021, 42: 100423.被引 43
  • Fast and Low Memory Cost Matrix Factorization: Algorithm, Analysis, and Case Study
    Yan, Yan*; Tan, Mingkui; Tsang, Ivor W.; Yang, Yi; Shi, Qinfeng; Zhang, ChengqiIEEE Transactions on Knowledge and Data Engineering, 2020, 32(2): 288-301.被引 3
  • Computing a nonnegative matrix factorization -- provably
    Sanjeev, Arora; Rong, Ge; Ravindran, Kannan; Ankur, Moitrathe 44th symposium, 2012-05-19 To 2012-05-22.被引 60
查看全部 →

发表信息

论文状态
公开发表
期刊
Journal of Physics A: Mathematical and Theoretical
发表日期
2024-02-23
卷期
第 57 卷, 第 8 期
页码
085002
DOI
10.1088/1751-8121/ad2299
语言
ENGLISH
成果收录
SCIE
资助标注
PNRR-PE-AI FAIR project - NextGeneration EU program

作者

  • C

    Camilli, Francesco (通讯) (第一作者)

  • M

    Mezard, Marc

登录解锁更多功能

  • 收藏这篇,随时在“我的 · 收藏”找回
  • 认领你的成果,生成推广素材
  • 关注作者获取最新成果
  • 加入项目组、协同写作
注册 / 登录

产品服务

  • 科研之友
  • 创新城
  • 科创云

服务支持

  • 帮助中心
  • 隐私政策
  • 服务条款

联系方式

在线客服:【】
客服热线:400-1616-289
电子邮箱:support@scholarmate.com

关注或下载科研之友

微信二维码
微信公众号
客户端下载二维码
下载客户端
科研成果科研人员科研机构科研动态爱瑞思软件

©2026 深圳市科研之友网络服务有限公司

公安备案图标粤公网安备 44030502000213
粤ICP备 16046710 号粤B2-20110417