A Stochastic Composite Gradient Method with Incremental Variance Reduction

被引:0
|
作者
Zhang, Junyu [1 ]
Xiao, Lin [2 ]
机构
[1] Univ Minnesota, Minneapolis, MN 55455 USA
[2] Microsoft Res, Redmond, WA 98052 USA
关键词
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
We consider the problem of minimizing the composition of a smooth (nonconvex) function and a smooth vector mapping, where the inner mapping is in the form of an expectation over some random variable or a finite sum. We propose a stochastic composite gradient method that employs an incremental variance-reduced estimator for both the inner vector mapping and its Jacobian. We show that this method achieves the same orders of complexity as the best known first-order methods for minimizing expected-value and finite-sum nonconvex functions, despite the additional outer composition which renders the composite gradient estimator biased. This finding enables a much broader range of applications in machine learning to benefit from the low complexity of incremental variance-reduction methods.
引用
收藏
页数:11
相关论文
共 50 条
  • [1] PROXIMAL STOCHASTIC GRADIENT METHOD WITH PROGRESSIVE VARIANCE REDUCTION
    Xiao, Lin
    Zhang, Tong
    [J]. SIAM JOURNAL ON OPTIMIZATION, 2014, 24 (04) : 2057 - 2075
  • [2] Stochastic gradient descent with variance reduction technique
    Zhang, Jinjing
    Hu, Fei
    Xu, Xiaofei
    Li, Li
    [J]. WEB INTELLIGENCE, 2018, 16 (03) : 187 - 194
  • [3] Stochastic Gradient Langevin Dynamics with Variance Reduction
    Huang, Zhishen
    Becker, Stephen
    [J]. 2021 INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS (IJCNN), 2021,
  • [4] Variance Reduction in Stochastic Gradient Langevin Dynamics
    Dubey, Avinava
    Reddi, Sashank J.
    Poczos, Barnabas
    Smola, Alexander J.
    Xing, Eric P.
    [J]. ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 29 (NIPS 2016), 2016, 29
  • [5] Stochastic Conjugate Gradient Algorithm With Variance Reduction
    Jin, Xiao-Bo
    Zhang, Xu-Yao
    Huang, Kaizhu
    Geng, Guang-Gang
    [J]. IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS, 2019, 30 (05) : 1360 - 1369
  • [6] A Composite Randomized Incremental Gradient Method
    Zhang, Junyu
    Xiao, Lin
    [J]. INTERNATIONAL CONFERENCE ON MACHINE LEARNING, VOL 97, 2019, 97
  • [7] Distributed and asynchronous Stochastic Gradient Descent with variance reduction
    Ming, Yuewei
    Zhao, Yawei
    Wu, Chengkun
    Li, Kuan
    Yin, Jianping
    [J]. NEUROCOMPUTING, 2018, 281 : 27 - 36
  • [8] On the Theory of Variance Reduction for Stochastic Gradient Monte Carlo
    Chatterji, Niladri S.
    Flammarion, Nicolas
    Ma, Yi-An
    Bartlett, Peter L.
    Jordan, Michael I.
    [J]. INTERNATIONAL CONFERENCE ON MACHINE LEARNING, VOL 80, 2018, 80
  • [9] Stochastic variable metric proximal gradient with variance reduction for non-convex composite optimization
    Gersende Fort
    Eric Moulines
    [J]. Statistics and Computing, 2023, 33 (3)
  • [10] Stochastic variable metric proximal gradient with variance reduction for non-convex composite optimization
    Fort, Gersende
    Moulines, Eric
    [J]. STATISTICS AND COMPUTING, 2023, 33 (03)