Summary

Proceedings of the 2013 International Symposium on Nonlinear Theory and its Applications

2013

Session Number:C1L-D

Session:

Number:386

Back Propagation Learning of Neural Networks with Replicated Neurons

Naohiro Shibuya,  Yuta Yokoyama,  Chihiro Ikuta,  Yoko Uwate,  Yoshifumi Nishio,  

pp.386-389

Publication Date:

Online ISSN:2188-5079

DOI:10.15248/proc.2.386

PDF download (1.4MB)

The human brain is able to process the complex information. One of the reason is that the cerebellum has a particular function. This function is that the cerebellum copies information in the cerebrum. We focus on the function of the cerebellum.
In this study, we apply such function to the artificial neural network operating the Back Propagation (BP). We actualize the function of the cerebellum by some processing. We consider that the number of one neuron is doubled at arbitrary timing. The weight parameters of one neuron are taken over in the doubled neurons. The method of taking over the weight parameters is that the weight parameters are carved up and sort weight parameters into each neuron. When the weight parameters are taken over, the different errors are added to each weight parameter. The doubled neurons learn with small learning rate. We confirm that the learning performance of the proposed network is better than the other networks.

[1] T. Poggio and F. Girosi, “Networks for approximation and learning”, Proc. IEEE, vol.78, no.9, pp.1481-1497, 1990.

[2] K. Hornik , M. Stinchcombe and H. White, “Multilayer feedforward networks are universal approximators”, Neural Networks, vol.2, pp.359-366, 1989.

[3] Y. Uwate and Y. Nishio, “Durability of Affordable Neural Networks against Damages”, International Joint Conference on Neural Networks (IJCNN'06), pp.8365-8370, July, 2006.

[4] D.E. Rumelhart, G.E. Hinton, and R.J. Williams, “Learning representations by back-propagating errors”, Nature, vol.323-9, pp.533-536, 1986.

[5] Takeshi Agui, Hiroshi Nagahashi, Hiroki Takahashi, “Neural Program”, Shokoudou corp, pp.11-18, 20-42, 1995.

[6] Jain. Rarnesh C, “A robust backpropagation learning algorithm for function approximation”, Journals, vol.5, pp.467-479, May, 1994.

[7] Horng-Lin Shieh, Ying-Kuei Yang, Po-Lun Chang, Jin-Tsong Jeng, “Robust neural-fuzzy method for function approximation”, Expert Systems with Applications, vol.36, pp.6903-6913, April, 2009.