Faster training of very deep networks via p-norm gates
Pham, Trang, Tran, Truyen, Phung, Dinh and Venkatesh, Svetha 2016, Faster training of very deep networks via p-norm gates, in ICPR 2016: Proceedings of the 23rd International Conference on Pattern Recognition, IEEE, Piscataway, N.J., pp. 3542-3547, doi: 10.1109/ICPR.2016.7900183.
ICPR 2016: Proceedings of the 23rd International Conference on Pattern Recognition
Publication date
2016
Conference series
Pattern Recognition International Conference
Start page
3542
End page
3547
Total pages
6
Publisher
IEEE
Place of publication
Piscataway, N.J.
Summary
A major contributing factor to the recent advances in deep neural networks is structural units that let sensory information and gradients to propagate easily. Gating is one such structure that acts as a flow control. Gates are employed in many recent state-of-the-art recurrent models such as LSTM and GRU, and feedforward models such as Residual Nets and Highway Networks. This enables learning in very deep networks with hundred layers and helps achieve record-breaking results in vision (e.g., ImageNet with Residual Nets) and NLP (e.g., machine translation with GRU). However, there is limited work in analysing the role of gating in the learning process. In this paper, we propose a flexible p-norm gating scheme, which allows usercontrollable flow and as a consequence, improve the learning speed. This scheme subsumes other existing gating schemes, including those in GRU, Highway Networks and Residual Nets as special cases. Experiments on large sequence and vector datasets demonstrate that the proposed gating scheme helps improve the learning speed significantly without extra overhead.
Unless expressly stated otherwise, the copyright for items in DRO is owned by the author, with all rights reserved.
Every reasonable effort has been made to ensure that permission has been obtained for items included in DRO.
If you believe that your rights have been infringed by this repository, please contact drosupport@deakin.edu.au.
Every reasonable effort has been made to ensure that permission has been obtained for items included in DRO. If you believe that your rights have been infringed by this repository, please contact drosupport@deakin.edu.au.