A parallel variable memory BFGS training algorithm

被引：0
|
作者：
Mc Loone, S [1 ]
机构：
[1] Queens Univ Belfast, Sch Elect & Elect Engn, Intelligent Syst & Control Res Grp, Belfast BT9 5AH, Antrim, North Ireland
来源：
ALGORITHMS AND ARCHITECTURES FOR REAL-TIME CONTROL 2000 | 2000年
关键词：
neural networks; parallel algorithms; training; second-order;
D O I：
暂无
中图分类号：
TP [自动化技术、计算机技术];
学科分类号：
0812 ;
摘要：
This paper considers the parallel implementation of a novel variable memory quasi-newton neural network training algorithm recently developed by the author. Unlike existing training methods this new technique is able to optimize performance in relation to available memory. Numerically it has equivalent properties to Full Memory BFGS optimization (FM) when there are no restrictions on memory and to FM with periodic reset when memory is limited. Parallel implementations of both the Full and Variable Memory BFGS algorithms are outlined and performance results presented for a PVM target architecture. Copyright (C) 2000 IFAC.
引用
收藏
页码：129 / 134
页数：4
相关论文
共 50 条
  • [21] Diagonal BFGS updates and applications to the limited memory BFGS method
    Li, Donghui
    Wang, Xiaozhou
    Huang, Jiajian
    COMPUTATIONAL OPTIMIZATION AND APPLICATIONS, 2022, 81 (03) : 829 - 856
  • [22] Diagonal BFGS updates and applications to the limited memory BFGS method
    Donghui Li
    Xiaozhou Wang
    Jiajian Huang
    Computational Optimization and Applications, 2022, 81 : 829 - 856
  • [23] A Riemannian Limited-Memory BFGS Algorithm for Computing the Matrix Geometric Mean
    Yuan, Xinru
    Huang, Wen
    Absil, P-A.
    Gallivan, K. A.
    INTERNATIONAL CONFERENCE ON COMPUTATIONAL SCIENCE 2016 (ICCS 2016), 2016, 80 : 2147 - 2157
  • [24] Parallel SSLE algorithm with parallel variable distribution for constrained optimization
    Zheng, FY
    He, GP
    CCCT 2003, VOL6, PROCEEDINGS: COMPUTER, COMMUNICATION AND CONTROL TECHNOLOGIES: III, 2003, : 349 - 354
  • [25] Regularized Stochastic BFGS algorithm
    Mokhtari, Aryan
    Ribeiro, Alejandro
    2013 IEEE GLOBAL CONFERENCE ON SIGNAL AND INFORMATION PROCESSING (GLOBALSIP), 2013, : 1109 - 1112
  • [26] Gradient trust region algorithm with limited memory BFGS update for nonsmooth convex minimization
    Gonglin Yuan
    Zengxin Wei
    Zhongxing Wang
    Computational Optimization and Applications, 2013, 54 : 45 - 64
  • [27] Gradient trust region algorithm with limited memory BFGS update for nonsmooth convex minimization
    Yuan, Gonglin
    Wei, Zengxin
    Wang, Zhongxing
    COMPUTATIONAL OPTIMIZATION AND APPLICATIONS, 2013, 54 (01) : 45 - 64
  • [28] Darwinian evolution in parallel universes: A parallel genetic algorithm for variable selection
    Zhu, Mu
    Chipman, Hugh A.
    TECHNOMETRICS, 2006, 48 (04) : 491 - 502
  • [29] Modified subspace limited memory BFGS algorithm for large-scale bound constrained optimization
    Xiao, Yunhai
    Zhang, Hongchuan
    JOURNAL OF COMPUTATIONAL AND APPLIED MATHEMATICS, 2008, 222 (02) : 429 - 439
  • [30] BFGS with update skipping and varying memory
    Kolda, TG
    O'Leary, DP
    Nazareth, L
    SIAM JOURNAL ON OPTIMIZATION, 1998, 8 (04) : 1060 - 1083
12345