SPSA for Layer-Wise Training of Deep Networks

被引:2
|
作者
Wulff, Benjamin [1 ,2 ]
Schuecker, Jannis [1 ,2 ]
Bauckhage, Christian [1 ,2 ,3 ]
机构
[1] Fraunhofer Ctr Machine Learning, St Augustin, Germany
[2] Fraunhofer IAIS, St Augustin, Germany
[3] Univ Bonn, B IT, Bonn, Germany
来源
ARTIFICIAL NEURAL NETWORKS AND MACHINE LEARNING - ICANN 2018, PT III | 2018年 / 11141卷
关键词
D O I
10.1007/978-3-030-01424-7_55
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Concerned with neural learning without backpropagation, we investigate variants of the simultaneous perturbation stochastic approximation (SPSA) algorithm. Experimental results suggest that these allow for the successful training of deep feed-forward neural networks using forward passes only. In particular, we find that SPSA-based algorithms which update network parameters in a layer-wise manner are superior to variants which update all weights simultaneously.
引用
收藏
页码:564 / 573
页数:10
相关论文
共 50 条
  • [1] Voice Conversion Using Deep Neural Networks With Layer-Wise Generative Training
    Chen, Ling-Hui
    Ling, Zhen-Hua
    Liu, Li-Juan
    Dai, Li-Rong
    IEEE-ACM TRANSACTIONS ON AUDIO SPEECH AND LANGUAGE PROCESSING, 2014, 22 (12) : 1859 - 1872
  • [2] Supervised Greedy Layer-Wise Training for Deep Convolutional Networks with Small Datasets
    Rueda-Plata, Diego
    Ramos-Pollan, Raul
    Gonzalez, Fabio A.
    COMPUTATIONAL COLLECTIVE INTELLIGENCE (ICCCI 2015), PT I, 2015, 9329 : 275 - 284
  • [3] Layer-Wise Compressive Training for Convolutional Neural Networks
    Grimaldi, Matteo
    Tenace, Valerio
    Calimera, Andrea
    FUTURE INTERNET, 2019, 11 (01)
  • [4] Layer-Wise Weight Decay for Deep Neural Networks
    Ishii, Masato
    Sato, Atsushi
    IMAGE AND VIDEO TECHNOLOGY (PSIVT 2017), 2018, 10749 : 276 - 289
  • [5] Stochastic Layer-Wise Precision in Deep Neural Networks
    Lacey, Griffin
    Taylor, Graham W.
    Areibi, Shawki
    UNCERTAINTY IN ARTIFICIAL INTELLIGENCE, 2018, : 663 - 672
  • [6] Unsupervised Layer-Wise Model Selection in Deep Neural Networks
    Ludovic, Arnold
    Helene, Paugam-Moisy
    Michele, Sebag
    ECAI 2010 - 19TH EUROPEAN CONFERENCE ON ARTIFICIAL INTELLIGENCE, 2010, 215 : 915 - 920
  • [7] Layer-Wise Training to Create Efficient Convolutional Neural Networks
    Zeng, Linghua
    Tian, Xinmei
    NEURAL INFORMATION PROCESSING (ICONIP 2017), PT II, 2017, 10635 : 631 - 641
  • [8] Collaborative Layer-Wise Discriminative Learning in Deep Neural Networks
    Jin, Xiaojie
    Chen, Yunpeng
    Dong, Jian
    Feng, Jiashi
    Yan, Shuicheng
    COMPUTER VISION - ECCV 2016, PT VII, 2016, 9911 : 733 - 749
  • [9] Multithreaded Layer-wise Training of Sparse Deep Neural Networks using Compressed Sparse Column
    Mofrad, Mohammad Hasanzadeh
    Melhem, Rami
    Ahmad, Yousuf
    Hammoud, Mohammad
    2019 IEEE HIGH PERFORMANCE EXTREME COMPUTING CONFERENCE (HPEC), 2019,
  • [10] GREEDY LAYER-WISE TRAINING OF LONG SHORT TERM MEMORY NETWORKS
    Xu, Kaisheng
    Shen, Xu
    Yao, Ting
    Tian, Xinmei
    Mei, Tao
    2018 IEEE INTERNATIONAL CONFERENCE ON MULTIMEDIA & EXPO WORKSHOPS (ICMEW 2018), 2018,