A Novel Approach for Efficient Training of Deep Neural Networks
Deep Neural Network training algorithms consumes long training time, especially when the number of hidden layers and nodes is large. Matrix multiplication is the key operation carried out at every node of each layer for several hundreds of thousands of times during the training of Deep Neural Networ...
Saved in:
Main Authors: | , |
---|---|
Format: | EJournal Article |
Published: |
Institute of Advanced Engineering and Science,
2018-09-01.
|
Subjects: | |
Online Access: | Get fulltext |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
LEADER | 02367 am a22003013u 4500 | ||
---|---|---|---|
001 | ijeecs13206_9135 | ||
042 | |a dc | ||
100 | 1 | 0 | |a Dharmajee Rao, D.T.V. |e author |
100 | 1 | 0 | |e contributor |
700 | 1 | 0 | |a Ramana, K.V. |e author |
245 | 0 | 0 | |a A Novel Approach for Efficient Training of Deep Neural Networks |
260 | |b Institute of Advanced Engineering and Science, |c 2018-09-01. | ||
500 | |a https://ijeecs.iaescore.com/index.php/IJEECS/article/view/13206 | ||
520 | |a Deep Neural Network training algorithms consumes long training time, especially when the number of hidden layers and nodes is large. Matrix multiplication is the key operation carried out at every node of each layer for several hundreds of thousands of times during the training of Deep Neural Network. Blocking is a well-proven optimization technique to improve the performance of matrix multiplication. Blocked Matrix multiplication algorithms can easily be parallelized to accelerate the performance further. This paper proposes a novel approach of implementing Parallel Blocked Matrix multiplication algorithms to reduce the long training time. The proposed approach was implemented using a parallel programming model OpenMP with collapse() clause for the multiplication of input and weight matrices of Backpropagation and Boltzmann Machine Algorithms for training Deep Neural Network and tested on multi-core processor system. Experimental results showed that the proposed approach achieved approximately two times speedup than classic algorithms. | ||
540 | |a Copyright (c) 2018 Institute of Advanced Engineering and Science | ||
540 | |a http://creativecommons.org/licenses/by-nc/4.0 | ||
546 | |a eng | ||
690 | |||
690 | |a Deep Neural Network; Parallel Blocked Matrix multiplication; Backpropagation and Boltzmann Machine algorithms; OpenMP; Multi-core processor system | ||
655 | 7 | |a info:eu-repo/semantics/article |2 local | |
655 | 7 | |a info:eu-repo/semantics/publishedVersion |2 local | |
655 | 7 | |2 local | |
786 | 0 | |n Indonesian Journal of Electrical Engineering and Computer Science; Vol 11, No 3: September 2018; 954-961 | |
786 | 0 | |n 2502-4760 | |
786 | 0 | |n 2502-4752 | |
786 | 0 | |n 10.11591/ijeecs.v11.i3 | |
787 | 0 | |n https://ijeecs.iaescore.com/index.php/IJEECS/article/view/13206/9135 | |
856 | 4 | 1 | |u https://ijeecs.iaescore.com/index.php/IJEECS/article/view/13206/9135 |z Get fulltext |