A Circuit-Based Neural Network with Hybrid Learning of Backpropagationand Random Weight Change Algorithms

Collection with item attached
2017
Item details URL
http://open-repository.kisti.re.kr/cube/handle/open_repository/473489.do
DOI
10.3390/s17010016
Title
A Circuit-Based Neural Network with Hybrid Learning of Backpropagationand Random Weight Change Algorithms
Description
This research was supported in part by the U.S. Air Force GrantFA9550-13-1-0136, the Basic Science Research Program through theNational Research Foundation of Korea (NRF) funded by the Ministry ofEducation (NRF-2016R1A2B4015514, NRF-2015H1D3A1062316) and the BrainKorea 21 PLUS Project, National Research Foundation of Korea.
abstract
Ahybrid learning method of a software-based backpropagation learning and a hardware-based RWC learning is proposed for the development of circuit-based neural networks. The backpropagation is known as one of the most efficient learning algorithms. A weak point is that its hardware implementation is extremely difficult. The RWC algorithm, which is very easy to implement with respect to its hardware circuits, takes too many iterations for learning. The proposed learning algorithm is a hybrid one of these two. The main learning is performed with a software version of the BP algorithm, firstly, and then, learned weights are transplanted on a hardware version of a neural circuit. At the time of the weight transplantation, a significant amount of output error would occur due to the characteristic difference between the software and the hardware. In the proposed method, such error is reduced via a complementary learning of the RWC algorithm, which is implemented in a simple hardware. The usefulness of the proposed hybrid learning system is verified via simulations upon several classical learning problems.
provenance
Made available in Cube on 2018-09-28T10:20:58Z (GMT). No. of bitstreams: 0
language
English
author
Yang, Changju
Kim, Hyongsuk
Adhikari, Shyam Prasad
Chua, Leon O.
accessioned
2018-09-28T10:20:58Z
available
2018-09-28T10:20:58Z
issued
2017
citation
SENSORS(17): 1
issn
1424-8220
uri
http://open-repository.kisti.re.kr/cube/handle/open_repository/473489.do
Funder
과학기술정보통신부
Funding Program
개인기초연구(미래부)
Project ID
1711054247
Jurisdiction
Rep.of Korea
Project Name
Development of a Convolutional Deep Learning Neural Networks for a Circuit Implementation of Human Intelligence
rights
openAccess
subject
software-based learning
circuit-based learning
complementary learning
backpropagation
RWC
type
article


Files in This Item

There are no attached files.