BP Network Optimization Based on Improved Genetic Algorithm

Abstract:

In view of the virtue and shortage of genetic algorithm and BP network, this paper proposes a new BP network training method based on improved genetic algorithm (IGA-BP). This algorithm uses hierarchical code, adaptive crossover and mutation, pruning similar chromosomes, dynamic supply new chromosomes and other operations, so the network structure and weight are optimized at the same time and the "premature" phenomenon is avoided. The simulation results show that the IGA-BP network architecture is simple, the convergence rate is quick, and has good approximation and generalization ability.

• Periodical: Advanced Materials Research (Volumes 532-533)
• Pages: 1757-1763
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.532-533.1757
• Online since: June 2012
• Authors: Wei Wei Sun, Yun Fei Yao, Chun Sheng Wang, Ye Gang Hu
• Keywords: Adaptive, BP Network, Genetic Algorithm (GA), Hierarchical Code, Pruning

References

[1] Yan Pingfan, Zhang Changshui. Artificial neural networks and simulation evolutionary computation. Beijing, Tsinghua University Press, (2005).

[2] Gary G. Yen, Haiming Lu. Hierarch genetic algorithm based neural network design. IEEE Symposium on Combination of Evolutionary Computation and Neural Networks, 168-175, (2000).

DOI: 10.1109/ecnn.2000.886232

[3] Jin Chaohong, Wu Hansong, Li Lamei. A neural networks training algorithm based on adaptive genetic algorithm. Control and Automation, 21(10): 49-51, (2005).

[4] Xu lu, Tu Chengyu. A new improved genetic algorithm and its property analysis. Acta Electronica Sinica, 29 (7): 902-904, (2001).

[5] Wei Gao. Study on new evolutionary neural network. Proceedings of the 2003 International Conference on Machine Learning and Cybernetics, 1287-1292.

[6] Frank H. F. Leung, H. K. Lam, S. H. Ling et al. Tuning of the structure and parameters of a neural network using an improved genetic algorithm. IEEE Trans on Neural Networks, 14 (1): 79-81, (2003).

DOI: 10.1109/tnn.2002.804317

[7] G A Rovithakis, I Chalkiadakis, M E Zervakis. High-order neural network structure selection for function approximation applications using genetic algorithms. IEEE Trans Systems, Man and Cybernetics, 1: 153-159, (2003).

DOI: 10.1109/tsmcb.2003.811767

[8] Srinvas M, Patnaik L M. Adaptive Probabilities of Crossover and Mutation in Genetic Algorithm. IEEE Trans on Systems, Man and Cybernetics, 24 (4): 162-167, (1994).

DOI: 10.1109/21.286385

[9] Shiwei Yu, Kejun Zhu, Fengqin Diao. A dynamic all parameters adaptive BP neural networks model and its application on oil reservoir prediction. Applied Mathematics and Computaion, 195: 66-75, (2008).

DOI: 10.1016/j.amc.2007.04.088