119 
(c) 1 (d) 
Figure 4: The behavior of squared error function. 
4. CONCLUSION 
In this paper we consider that the communication control in 
mutual connected network by reproductive and competitive 
radial basis function network. We propose RC-RBFN by 
applying the synaptic plasticity equation as the survival of the 
fittest learning to the RBFN. The CRBFN, which we first 
propose can faster learning by using the survived minimum, 
required input neurons. From the results of simulation, it is 
shown that the RC-RBFN can estimate the network 
parameters by using optimum number of neurons after 
eliminating the redundant neurons. 
The further problem is that we exponent the synaptic 
plasticity equation as survival of the fittest learning algorithm 
to apply other network problems. 
REFERENCE 
[1] J. Park, and I. W. Sandberg, "Universal approximation 
using radial basis function networks". Neural Computation, 
3, pp. 246-257, 1991. 
[2] K. Funahashi, "On the approximate realization of 
continuous mappings by neural networks". Neural Networks, 
2, pp. 148-153, 1989. 
[3] D. B. Fogel, " An information criterion for optimal neural 
network selection". IEEE Trans. Neural Networks, 2, pp. 
490-497, 1991. 
[4] Z. Wang, C. D. Massimo, M. T. Tham, and A. J. Morris, 
"A procedure for determining the topology for multilayer 
feedforward networks". Neural Networks, 7, pp. 291-300, 
1994. 
[5] B. Widrow, and R. Winter, " Neural nets for adaptive 
filtering and adaptive pattern recognition ". IEEE Computer, 
21, pp. 25-39, 1988. 
[6] H. Haken, “Synergetics - An Introduction: 
Nonequilibrium Phase Transitions and Self-Organization in 
Physics, Chemistry and Biology,” Springer-Verlag, 1978. 
[7] Y. Wong, “Clustering data by melting,” Neural 
Computation, 5, pp. 89-104, 1993.