A General Algorithm for Solving Rank-one Matrix Sensing

被引:0
|
作者
Qin, Lianke [1 ]
Song, Zhao [2 ]
Zhang, Ruizhe [3 ]
机构
[1] UC Santa Barbara, Santa Barbara, CA 93106 USA
[2] Adobe Res, San Francisco, CA USA
[3] Univ Calif Berkeley, Simons Inst, Berkeley, CA USA
关键词
INTERIOR-POINT METHOD;
D O I
暂无
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Matrix sensing has many real-world applications in science and engineering, such as system control, distance embedding, and computer vision. The goal of matrix sensing is to recover a matrix A(star) is an element of R-nxn, based on a sequence of measurements (u(i), b(i)) is an element of R(n)xR such that u(i)(inverted perpendicular) A(star)u(i) = b(i). Previous work (Zhong et al., 2015) focused on the scenario where matrix A. has a small rank, e.g. rank-k. Their analysis heavily relies on the RIP assumption, making it unclear how to generalize to high-rank matrices. In this paper, we relax that rank-k assumption and solve a much more general matrix sensing problem. Given an accuracy parameter delta is an element of (0, 1), we can compute A is an element of R-nxn in (O) over tilde (m(3/2)n(2)delta(-1)), such that vertical bar u(i)(inverted perpendicular) Au-i - b(i)vertical bar <= delta for all i is an element of [m]. We design an efficient algorithm with provable convergence guarantees using stochastic gradient descent for this problem.
引用
收藏
页数:27
相关论文
共 50 条
  • [21] Nonconvex Matrix Factorization from Rank-One Measurements
    Li, Yuanxin
    Ma, Cong
    Chen, Yuxin
    Chi, Yuejie
    22ND INTERNATIONAL CONFERENCE ON ARTIFICIAL INTELLIGENCE AND STATISTICS, VOL 89, 2019, 89
  • [22] ROP: MATRIX RECOVERY VIA RANK-ONE PROJECTIONS
    Cai, T. Tony
    Zhang, Anru
    ANNALS OF STATISTICS, 2015, 43 (01): : 102 - 138
  • [23] The bounds of the eigenvalues for rank-one modification of Hermitian matrix
    Cheng, Guanghui
    Song, Zhida
    Yang, Jianfeng
    Si, Jia
    NUMERICAL LINEAR ALGEBRA WITH APPLICATIONS, 2014, 21 (01) : 98 - 107
  • [24] Nonconvex Matrix Factorization From Rank-One Measurements
    Li, Yuanxin
    Ma, Cong
    Chen, Yuxin
    Chi, Yuejie
    IEEE TRANSACTIONS ON INFORMATION THEORY, 2021, 67 (03) : 1928 - 1950
  • [25] Solving Bernoulli Rank-One Bandits with Unimodal Thompson Sampling
    Trinh, Cindy
    Kaufmann, Emilie
    Vernade, Claire
    Combes, Richard
    ALGORITHMIC LEARNING THEORY, VOL 117, 2020, 117 : 862 - 889
  • [26] Quark Mass Matrix with a Structure of a Rank-One Matrix Plus a Unit Matrix
    Fusaoka, H.
    Koide, Y.
    Mathematical and Computer Modelling (Oxford), 1600, 21 (05):
  • [27] A Non-convex One-Pass Framework for Generalized Factorization Machine and Rank-One Matrix Sensing
    Lin, Ming
    Ye, Jieping
    ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 29 (NIPS 2016), 2016, 29
  • [28] QUARK MASS MATRIX WITH A STRUCTURE OF A RANK-ONE MATRIX PLUS A UNIT MATRIX
    FUSAOKA, H
    KOIDE, Y
    MODERN PHYSICS LETTERS A, 1995, 10 (04) : 289 - 294
  • [29] A Rank-One Tensor Updating Algorithm for Tensor Completion
    Yang, Yuning
    Feng, Yunlong
    Suykens, Johan A. K.
    IEEE SIGNAL PROCESSING LETTERS, 2015, 22 (10) : 1633 - 1637
  • [30] APPROXIMATELY SYMMETRICAL KM MATRIX AND RANK-ONE QUARK MASS MATRIX
    TANIMOTO, M
    MODERN PHYSICS LETTERS A, 1991, 6 (25) : 2309 - 2314