Neuron - synapse level problem decomposition method for cooperative coevolution of recurrent networks for time series prediction

Nand, Ravneil (2016) Neuron - synapse level problem decomposition method for cooperative coevolution of recurrent networks for time series prediction. [Conference Proceedings]

PDF - Submitted Version Restricted to Repository staff only Download (360kB) | Request a copy

Abstract

The decomposition of a particular problem can become a tiresome task if little connections between the elements is needed. In cooperative coevolution of recurrent networks, synapse and neuron level are the two noteworthy problem decomposition methods. Through combination of both of the problem decomposition methods, the individual problem decomposition methods can share its strengths to solve the problem at hand better. In this paper, a recently proposed problem decomposition method known as Neuron-Synapse problem decomposition method is modified for Elman recurrent neural networks. The results reveal that the proposed method has got better results in selected datasets when compared to standalone methods. The results are better in some cases for proposed method when compared to other approaches from the literature.

Item Type:	Conference Proceedings
Subjects:	Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Divisions:	Faculty of Science, Technology and Environment (FSTE) > School of Computing, Information and Mathematical Sciences
Depositing User:	Ravneil Nand
Date Deposited:	05 Dec 2016 23:13
Last Modified:	08 Jul 2019 22:29
URI:	https://repository.usp.ac.fj/id/eprint/9187

Actions (login required)

View Item