Fast Monte-Carlo algorithms for finding low-rank approximations

被引:306
|
作者
Frieze, A [1 ]
Kannan, R
Vempala, S
机构
[1] Carnegie Mellon Univ, Pittsburgh, PA 15213 USA
[2] Yale Univ, New Haven, CT USA
[3] MIT, Cambridge, MA 02139 USA
关键词
algorithms; theory; matrix algorithms; sampling; low-rank approximation;
D O I
10.1145/1039488.1039494
中图分类号
TP3 [计算技术、计算机技术];
学科分类号
0812 ;
摘要
We consider the problem of approximating a given m x n matrix A by another matrix of specified rank k, which is smaller than in and n. The Singular Value Decomposition (SVD) can be used to find the "best" such approximation. However, it takes time polynomial in m, n which is prohibitive for some modern applications. In this article, we develop an algorithm that is qualitatively faster, provided we may sample the entries of the matrix in accordance with a natural probability distribution. In many applications, such sampling can be done efficiently. Our main result is a randomized algorithm to find the description of a matrix D* of rank at most k so that [GRAPHICS] holds with probability at least 1 - delta (where parallel to.parallel to(F) is the Frobenius norm). The algorithm takes time polynomial in k, 1/epsilon, log(1/delta) only and is independent of m and n. In particular, this implies that in constant time, it can be determined if a given matrix of arbitrary size has a good low-rank approximation.
引用
收藏
页码:1025 / 1041
页数:17
相关论文
共 50 条
  • [41] MONTE-CARLO ALGORITHMS WITH ABSORBING MARKOV-CHAINS - FAST LOCAL ALGORITHMS FOR SLOW DYNAMICS
    NOVOTNY, MA
    PHYSICAL REVIEW LETTERS, 1995, 74 (01) : 1 - 5
  • [42] DYNAMIC ANALYSIS OF LOW-TEMPERATURE MONTE-CARLO CLUSTER ALGORITHMS
    MARTINELLI, F
    JOURNAL OF STATISTICAL PHYSICS, 1992, 66 (5-6) : 1245 - 1276
  • [43] Correction to: ALORA: Affine Low-Rank Approximations
    Alan Ayala
    Xavier Claeys
    Laura Grigori
    Journal of Scientific Computing, 2019, 80 (3) : 1997 - 1997
  • [44] Two Rank Approximations for Low-Rank Based Subspace Clustering
    Xu, Fei
    Peng, Chong
    Hu, Yunhong
    He, Guoping
    2017 10TH INTERNATIONAL CONGRESS ON IMAGE AND SIGNAL PROCESSING, BIOMEDICAL ENGINEERING AND INFORMATICS (CISP-BMEI), 2017,
  • [45] Learning-Based Low-Rank Approximations
    Indyk, Piotr
    Vakilian, Ali
    Yuan, Yang
    ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 32 (NIPS 2019), 2019, 32
  • [46] Robust low-rank data matrix approximations
    XingDong Feng
    XuMing He
    Science China Mathematics, 2017, 60 : 189 - 200
  • [47] Parametric PDEs: sparse or low-rank approximations?
    Bachmayr, Markus
    Cohen, Albert
    Dahmen, Wolfgang
    IMA JOURNAL OF NUMERICAL ANALYSIS, 2018, 38 (04) : 1661 - 1708
  • [48] Robust low-rank data matrix approximations
    Feng XingDong
    He XuMing
    SCIENCE CHINA-MATHEMATICS, 2017, 60 (02) : 189 - 200
  • [49] Connectome Smoothing via Low-Rank Approximations
    Tang, Runze
    Ketcha, Michael
    Badea, Alexandra
    Calabrese, Evan D.
    Margulies, Daniel S.
    Vogelstein, Joshua T.
    Priebe, Carey E.
    Sussman, Daniel L.
    IEEE TRANSACTIONS ON MEDICAL IMAGING, 2019, 38 (06) : 1446 - 1456
  • [50] Low-rank approximations of nonseparable panel models
    Fernandez-Val, Ivan
    Freeman, Hugo
    Weidner, Martin
    ECONOMETRICS JOURNAL, 2021, 24 (02): : C40 - C77