Split Bregman method for large scale fused Lasso

被引:75
作者
Ye, Gui-Bo [1 ]
Xie, Xiaohui [1 ]
机构
[1] Univ Calif Irvine, Sch Informat & Comp Sci, Irvine, CA 92697 USA
基金
美国国家科学基金会;
关键词
Fused Lasso; Bregman iteration; l(1)-norm; Fused Lasso support vector classifier; IMAGE-RESTORATION; REGULARIZATION; RECONSTRUCTION; ALGORITHM; ITERATION;
D O I
10.1016/j.csda.2010.10.021
中图分类号
TP39 [计算机的应用];
学科分类号
081203 ; 0835 ;
摘要
Ordering of regression or classification coefficients occurs in many real-world applications. Fused Lasso exploits this ordering by explicitly regularizing the differences between neighboring coefficients through an l(1) norm regularizer. However, due to nonseparability and nonsmoothness of the regularization term, solving the fused Lasso problem is computationally demanding. Existing solvers can only deal with problems of small or medium size, or a special case of the fused Lasso problem in which the predictor matrix is the identity matrix. In this paper, we propose an iterative algorithm based on the split Bregman method to solve a class of large-scale fused Lasso problems, including a generalized fused Lasso and a fused Lasso support vector classifier. We derive our algorithm using an augmented Lagrangian method and prove its convergence properties. The performance of our method is tested on both artificial data and real-world applications including proteomic data from mass spectrometry and genomic data from array comparative genomic hybridization (array CGH). We demonstrate that our method is many times faster than the existing solvers, and show that it is especially efficient for large p, small n problems, where p is the number of variables and n is the number of samples. (C) 2010 Elsevier B.V. All rights reserved.
引用
收藏
页码:1552 / 1569
页数:18
相关论文
共 50 条
  • [1] Augmented Lagrangian Method, Dual Methods, and Split Bregman Iteration for ROF, Vectorial TV, and High Order Models
    Wu, Chunlin
    Tai, Xue-Cheng
    SIAM JOURNAL ON IMAGING SCIENCES, 2010, 3 (03): : 300 - 339
  • [2] Geometric Applications of the Split Bregman Method: Segmentation and Surface Reconstruction
    Goldstein, Tom
    Bresson, Xavier
    Osher, Stanley
    JOURNAL OF SCIENTIFIC COMPUTING, 2010, 45 (1-3) : 272 - 293
  • [3] Large-Scale Structured Sparsity via Parallel Fused Lasso on Multiple GPUs
    Lee, Taehoon
    Won, Joong-Ho
    Lim, Johan
    Yoon, Sungroh
    JOURNAL OF COMPUTATIONAL AND GRAPHICAL STATISTICS, 2017, 26 (04) : 851 - 864
  • [4] A SPLIT BREGMAN METHOD FOR LINEAR SPECTRAL UNMIXING
    Liu, Jianjun
    Wu, Zebin
    Wei, Zhihui
    Sun, Le
    2012 4TH WORKSHOP ON HYPERSPECTRAL IMAGE AND SIGNAL PROCESSING (WHISPERS), 2012,
  • [5] LATENT FUSED LASSO
    Feng, Yining
    Selesnick, Ivan
    2020 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH, AND SIGNAL PROCESSING, 2020, : 5969 - 5973
  • [6] A fast splitting method for efficient Split Bregman iterations
    Lazzaro, D.
    Piccolomini, E. Loli
    Zama, F.
    APPLIED MATHEMATICS AND COMPUTATION, 2019, 357 : 139 - 146
  • [7] Fast and accurate compressed sensing model in magnetic resonance imaging with median filter and split Bregman method
    Yang, Yunyun
    Qin, Xuxu
    Wu, Boying
    IET IMAGE PROCESSING, 2019, 13 (01) : 1 - 8
  • [8] Spatially adaptive total variation deblurring with split Bregman technique
    Dodangeh, Mahdi
    Figueiredo, Isabel N.
    Goncalves, Gil
    IET IMAGE PROCESSING, 2018, 12 (06) : 948 - 958
  • [9] A Compact Neural Network for Fused Lasso Signal Approximator
    Mohammadi, Majid
    IEEE TRANSACTIONS ON CYBERNETICS, 2021, 51 (08) : 4327 - 4336
  • [10] Lp Regularization for Bioluminescence Tomography Based on the Split Bregman Method
    Hu, Yifang
    Liu, Jie
    Leng, Chengcai
    An, Yu
    Zhang, Shuang
    Wang, Kun
    MOLECULAR IMAGING AND BIOLOGY, 2016, 18 (06) : 830 - 837