Scaling Up Kernel SVM on Limited Resources: A Low-Rank Linearization Approach

被引:37
作者
Lan, Liang [1 ]
Wang, Zhuang [2 ,3 ]
Zhe, Shandian [4 ]
Cheng, Wei [5 ]
Wang, Jun [6 ]
Zhang, Kai [7 ]
机构
[1] Hong Kong Baptist Univ, Dept Comp Sci, Hong Kong, Peoples R China
[2] Siemens Corp Res, Princeton, NJ 08540 USA
[3] Facebook, Menlo Pk, CA 94025 USA
[4] Univ Utah, Sch Comp, Salt Lake City, UT USA
[5] NEC Labs Amer, Princeton, NJ 08550 USA
[6] East China Normal Univ, Sch Comp Sci & Software Engn, Shanghai 200062, Peoples R China
[7] Temple Univ, Dept Comp & Informat Sci, Philadelphia, PA 19122 USA
关键词
Large-scale learning; low-rank approximation; metric learning; support vector machine (SVM); NYSTROM METHOD; MATRIX;
D O I
10.1109/TNNLS.2018.2838140
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Kernel support vector machines (SVMs) deliver state-of-the-art results in many real-world nonlinear classification problems, but the computational cost can be quite demanding in order to maintain a large number of support vectors. Linear SVM, on the other hand, is highly scalable to large data but only suited for linearly separable problems. In this paper, we propose a novel approach called low-rank linearized SVMto scale up kernel SVM on limited resources. Our approach transforms a nonlinear SVM to a linear one via an approximate empirical kernel map computed from efficient kernel low-rank decompositions. We theoretically analyze the gap between the solutions of the approximate and optimal rank-k kernel map, which in turn provides guidance on the sampling scheme of the Nystrom approximation. Furthermore, we extend it to a semisupervised metric learning scenario in which partially labeled samples can be exploited to further improve the quality of the low-rank embedding. Our approach inherits rich representability of kernel SVM and high efficiency of linear SVM. Experimental results demonstrate that our approach is more robust and achieves a better tradeoff between model representability and scalability against state-of-the-art algorithms for large-scale SVMs.
引用
收藏
页码:369 / 378
页数:10
相关论文
共 45 条
  • [1] [Anonymous], 2012, Proceedings of the fifteenth International Conference on Artificial Intelligence and Statistics
  • [2] [Anonymous], P 29 INT C MACH LEAR
  • [3] [Anonymous], 2013, C LEARN THEOR
  • [4] [Anonymous], 2011, PROC 17 ACM SIGKDD I
  • [5] Bach F. R., 2005, Proceedings of the 22nd International Conference Machine Learning Society, DOI [10.1145/1102351.1102356, DOI 10.1145/1102351.1102356]
  • [6] Fourier Kernel Learning
    Bazavan, Eduard Gabriel
    Li, Fuxin
    Sminchisescu, Cristian
    [J]. COMPUTER VISION - ECCV 2012, PT II, 2012, 7573 : 459 - 473
  • [7] Bordes A, 2005, J MACH LEARN RES, V6, P1579
  • [8] Chang C.-J., 2010, J. Mach. Learn. Res., V11, P1471
  • [9] LIBSVM: A Library for Support Vector Machines
    Chang, Chih-Chung
    Lin, Chih-Jen
    [J]. ACM TRANSACTIONS ON INTELLIGENT SYSTEMS AND TECHNOLOGY, 2011, 2 (03)
  • [10] CORTES C, 1995, MACH LEARN, V20, P273, DOI 10.1023/A:1022627411411