Multi-objective evolutionary recurrent neural networks for system identification

Please use this identifier to cite or link to this item: https://doi.org/10.1109/CEC.2007.4424662

Title:	Multi-objective evolutionary recurrent neural networks for system identification
Authors:	Ang, J.H. Goh, C.K. Teoh, E.J. Mamun, A.A.
Issue Date:	2007
Citation:	Ang, J.H., Goh, C.K., Teoh, E.J., Mamun, A.A. (2007). Multi-objective evolutionary recurrent neural networks for system identification. 2007 IEEE Congress on Evolutionary Computation, CEC 2007 : 1586-1592. ScholarBank@NUS Repository. https://doi.org/10.1109/CEC.2007.4424662
Abstract:	This paper proposes a new multi-objective evolutionary approach for training recurrent neural networks (RNNs). The algorithm uses features of a variable length representation allowing easy adaptation of neural networks structures and a micro genetic algorithm (μGA) with an adaptive local search intensity scheme for local fine-tuning. In addition, a structural mutation (SM) operator for evolving the appropriate number of neurons for RNNs is used. Simulation results demonstrated the effectiveness of proposed method for system identification tasks. © 2007 IEEE.
Source Title:	2007 IEEE Congress on Evolutionary Computation, CEC 2007
URI:	http://scholarbank.nus.edu.sg/handle/10635/71050
ISBN:	1424413400
DOI:	10.1109/CEC.2007.4424662
Appears in Collections:	Staff Publications

There are no files associated with this item.

Check