Title :
Convergence analysis of chaotic dynamic neuron
Author :
Kim, Sang-Hee ; Park, Won-Woo
Author_Institution :
Sch. of Electron. Eng., Kumoh Nat. Inst. of Tech., South Korea
Abstract :
This paper investigated the dynamic characteristics of a chaotic dynamic neuron. The transient chaos of neuron could be beneficial to overcome the local minimum problem, but the chaotic response gives adverse effect on optimization problems in general. In this paper, we analyze the condition that produce the chaotic response, and present the convergence condition.
Keywords :
convergence of numerical methods; dynamic response; neural nets; optimisation; chaotic dynamic neuron; chaotic neural network; chaotic response; convergence analysis; optimization problems; transient chaos; traveling salesman problem; Biological system modeling; Chaos; Convergence; Differential equations; Network address translation; Neural networks; Neurofeedback; Neurons; Transient analysis; Traveling salesman problems;
Conference_Titel :
Neural Networks, 2003. Proceedings of the International Joint Conference on
Print_ISBN :
0-7803-7898-9
DOI :
10.1109/IJCNN.2003.1223802