A divide-and-conquer algorithm for binary matrix completion

被引:5
作者
Beckerleg, Melanie [1 ]
Thompson, Andrew [1 ]
机构
[1] Univ Oxford, Math Inst, Andrew Wiles Bldg,Woodstock Rd, Oxford OX2 6GG, England
基金
英国工程与自然科学研究理事会;
关键词
Binary matrix completion; Linear programming; Recommender systems; FACTORIZATION; PREDICTION; DISCOVERY;
D O I
10.1016/j.laa.2020.04.017
中图分类号
O29 [应用数学];
学科分类号
070104 ;
摘要
We propose a practical algorithm for low rank matrix completion for matrices with binary entries which obtains explicit binary factors and show it performs well at the recommender task on real world datasets. The algorithm, which we call TBMC (Tiling for Binary Matrix Completion), gives interpretable output in the form of binary factors which represent a decomposition of the matrix into tiles. Our approach extends a popular algorithm from the data mining community, PROXIMUS, to missing data, applying the same recursive partitioning approach. The algorithm relies upon rank-one approximations of incomplete binary matrices, and we propose a linear programming (LP) approach for solving this subproblem. We also prove a 2-approximation result for the LP approach which holds for any subsampling pattern, and show that TBMC exactly solves the rank-k prediction task for a underlying block-diagonal tiling structure with geometrically decreasing tile sizes, providing the ratio between successive tiles is less than 1/root 2. Numerical experiments show that TBMC outperforms existing methods on real datasets. (C) 2020 Elsevier Inc. All rights reserved.
引用
收藏
页码:113 / 133
页数:21
相关论文
共 50 条
  • [21] Distributed Matrix Completion
    Teflioudi, Christina
    Makari, Faraz
    Gemulla, Rainer
    [J]. 12TH IEEE INTERNATIONAL CONFERENCE ON DATA MINING (ICDM 2012), 2012, : 655 - 664
  • [22] Statistical Optimality of Divide and Conquer Kernel-based Functional Linear Regression
    Liu, Jiading
    Shi, Lei
    [J]. JOURNAL OF MACHINE LEARNING RESEARCH, 2024, 25 : 1 - 56
  • [23] System Based Monitoring of a Neuromusculoskeletal System Using Divide and Conquer Type Models
    Musselman, Marcus
    Gates, Deanna
    Djurdjanovic, Dragan
    [J]. 2017 IEEE AEROSPACE CONFERENCE, 2017,
  • [24] Color Image Recovery Using Low-Rank Quaternion Matrix Completion Algorithm
    Miao, Jifei
    Kou, Kit Ian
    [J]. IEEE TRANSACTIONS ON IMAGE PROCESSING, 2022, 31 : 190 - 201
  • [25] BMCMDA: a novel model for predicting human microbe-disease associations via binary matrix completion
    Jian-Yu Shi
    Hua Huang
    Yan-Ning Zhang
    Jiang-Bo Cao
    Siu-Ming Yiu
    [J]. BMC Bioinformatics, 19
  • [26] BMCMDA: a novel model for predicting human microbe-disease associations via binary matrix completion
    Shi, Jian-Yu
    Huang, Hua
    Zhang, Yan-Ning
    Cao, Jiang-Bo
    Yiu, Siu-Ming
    [J]. BMC BIOINFORMATICS, 2018, 19
  • [27] Spectral Geometric Matrix Completion
    Boyarski, Amit
    Vedula, Sanketh
    Bronstein, Alex
    [J]. MATHEMATICAL AND SCIENTIFIC MACHINE LEARNING, VOL 145, 2021, 145 : 172 - 196
  • [28] Orthogonal Inductive Matrix Completion
    Ledent, Antoine
    Alves, Rodrigo
    Kloft, Marius
    [J]. IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS, 2023, 34 (05) : 2259 - 2270
  • [29] Matrix completion and vector completion via robust subspace learning
    Liu, Zhe
    Hu, Zhanxuan
    Nie, Feiping
    [J]. NEUROCOMPUTING, 2018, 306 : 171 - 181
  • [30] Carbon price forecasting system based on error correction and divide-conquer strategies
    Niu, Xinsong
    Wang, Jianzhou
    Zhang, Lifang
    [J]. APPLIED SOFT COMPUTING, 2022, 118