A RBF Neural Network Learning Algorithm Based on NCPSO

Xue Lei,Sun Changyin,Mu Chaoxu,Huang Yiqing
2013-01-01
Abstract:A RBF (Radial Basis Function) neural network learning algorithm based on NCPSO (Niching Chaotic Mutation Particle Swarm Optimization) is proposed. Because of the niching method and chaotic mutation, NCPSO can be used to optimize the output weights of the RBF Neural Network. Niching method is introduced to improve the ability of global optimization. Chaotic mutation is mentioned to improve the solution. Compared with the RBF neural network learning algorithm based on the GA (Genetic Algorithm), simulation shows that the RBF neural network learning algorithm based on NCPSO mentioned in this paper has a lower error of tracking and higher speed.
What problem does this paper attempt to address?