Nonnegative Matrix Factorization Via Archetypal Analysis

被引:7
|
作者
Javadi, Hamid [1 ]
Montanari, Andrea [2 ]
机构
[1] Rice Univ, Dept Elect & Comp Engn, POB 1892, Houston, TX 77005 USA
[2] Stanford Univ, Dept Elect Engn & Stat, Stanford, CA 94305 USA
基金
美国国家科学基金会;
关键词
Dimensionality reduction; Matrix factorization; Separability; ALGORITHMS;
D O I
10.1080/01621459.2019.1594832
中图分类号
O21 [概率论与数理统计]; C8 [统计学];
学科分类号
020208 ; 070103 ; 0714 ;
摘要
Given a collection of data points, nonnegative matrix factorization (NMF) suggests expressing them as convex combinations of a small set of "archetypes" with nonnegative entries. This decomposition is unique only if the true archetypes are nonnegative and sufficiently sparse (or the weights are sufficiently sparse), a regime that is captured by the separability condition and its generalizations. In this article, we study an approach to NMF that can be traced back to the work of Cutler and Breiman [(1994), "Archetypal Analysis," Technometrics, 36, 338-347] and does not require the data to be separable, while providing a generally unique decomposition. We optimize a trade-off between two objectives: we minimize the distance of the data points from the convex envelope of the archetypes (which can be interpreted as an empirical risk), while also minimizing the distance of the archetypes from the convex envelope of the data (which can be interpreted as a data-dependent regularization). The archetypal analysis method of Cutler and Breiman is recovered as the limiting case in which the last term is given infinite weight. We introduce a "uniqueness condition" on the data which is necessary for identifiability. We prove that, under uniqueness (plus additional regularity conditions on the geometry of the archetypes), our estimator is robust. While our approach requires solving a nonconvex optimization problem, we find that standard optimization methods succeed in finding good solutions for both real and synthetic data. for this article are available online
引用
收藏
页码:896 / 907
页数:12
相关论文
共 50 条
  • [41] Bayesian mean-parameterized nonnegative binary matrix factorization
    Lumbreras, Alberto
    Filstroff, Louis
    Fevotte, Cedric
    DATA MINING AND KNOWLEDGE DISCOVERY, 2020, 34 (06) : 1898 - 1935
  • [42] Nonnegative Matrix Factorization and Its Application to Pattern Analysis and Text Mining
    Zurada, Jacek M.
    Ensari, Tolga
    Asl, Ehsan Hosseini
    Chorowski, Jan
    2013 FEDERATED CONFERENCE ON COMPUTER SCIENCE AND INFORMATION SYSTEMS (FEDCSIS), 2013, : 11 - 16
  • [43] Text mining using nonnegative matrix factorization and latent semantic analysis
    Ali Hassani
    Amir Iranmanesh
    Najme Mansouri
    Neural Computing and Applications, 2021, 33 : 13745 - 13766
  • [44] Text mining using nonnegative matrix factorization and latent semantic analysis
    Hassani, Ali
    Iranmanesh, Amir
    Mansouri, Najme
    NEURAL COMPUTING & APPLICATIONS, 2021, 33 (20) : 13745 - 13766
  • [45] Smooth Nonnegative Matrix Factorization for Unsupervised Audiovisual Document Structuring
    Essid, Slim
    Fevotte, Cedric
    IEEE TRANSACTIONS ON MULTIMEDIA, 2013, 15 (02) : 415 - 425
  • [46] Fast and Robust Recursive Algorithms for Separable Nonnegative Matrix Factorization
    Gillis, Nicolas
    Vavasis, Stephen A.
    IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2014, 36 (04) : 698 - 714
  • [47] Ellipsoidal Rounding for Nonnegative Matrix Factorization Under Noisy Separability
    Mizutani, Tomohiko
    JOURNAL OF MACHINE LEARNING RESEARCH, 2014, 15 : 1011 - 1039
  • [48] COMPUTING A NONNEGATIVE MATRIX FACTORIZATION-PROVABLY
    Arora, Sanjeev
    Ge, Rong
    Kannan, Ravi
    Moitra, Ankur
    SIAM JOURNAL ON COMPUTING, 2016, 45 (04) : 1582 - 1611
  • [49] Adaptive Kernel Graph Nonnegative Matrix Factorization
    Li, Rui-Yu
    Guo, Yu
    Zhang, Bin
    INFORMATION, 2023, 14 (04)
  • [50] MINIMAX LOWER BOUNDS FOR NONNEGATIVE MATRIX FACTORIZATION
    Alsan, Mine
    Liu, Zhaoqiang
    Tan, Vincent Y. F.
    2018 IEEE STATISTICAL SIGNAL PROCESSING WORKSHOP (SSP), 2018, : 363 - 367