A Train Algorithm of Multi-Aggregation Process Neural Networks Based on Chaos Genetic Optimization

DOI

10.2991/citcs.2012.147How to use a DOI?

Keywords

process neural networks, train algorithm, chaos genetic algorithm, mixed optimization strategy

Abstract

Aiming at the learning problem of multi-aggregation process neural networks, an optimization train method based on chaos genetic algorithm (CGA) with lowest mean square algorithm (LMS) is proposed in the paper. The learning problem of the network weight functions is converted to the training of basis expansion coefficients through adopting the function orthogonal basis expansion method. The lowest mean square error as the train objective function, the global optimal solution of network parameters is solved in feasible solution space using the chaos rail to traverse the search of CGA. The application in multi-variant dynamic signal identification proved that the algorithm proposed in the paper improved the training efficiency and stability greatly

Copyright

© 2012, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Yue-Wu Pang
AU  - Shao-Hua XU
PY  - 2012/11
DA  - 2012/11
TI  - A Train Algorithm of Multi-Aggregation Process Neural Networks Based on Chaos Genetic Optimization
BT  - Proceedings of the 2012 National Conference on Information Technology and Computer Science
PB  - Atlantis Press
SP  - 571
EP  - 574
SN  - 1951-6851
UR  - https://doi.org/10.2991/citcs.2012.147
DO  - 10.2991/citcs.2012.147
ID  - Pang2012/11
ER  -