A Novel Training Algorithm for BP Neural Network

摘要：

In this paper, a training algorithm of neural network based on QPSO is proposed. Via training neural network with the quantum particle swarm optimization, the optimized combination of weights and bias of the network is attained.The application of function approximation and data classification is performed using the algorithm in the paper.The experiment result shows that the algorithm has better convergent speed and better global convergent characteristic compared with traditional neural network training algorithm.

关键词： Neural Network Training QPSO Optimized Combination of Weights and Bias Global ConvergentCharacteristic Better Convergent Speed

作者: Fang Bao Yonghui Pan Wenbo Xu

作者单位: School of Information Technology, Southern Yangtze University No.1800, Lihudadao, Wuxi Jiangsu 21412 School of Information Technology, Southern Yangtze University No.1800, Lihudadao, Wuxi Jiangsu 21412

会议类型: 国际会议

会议名称: 2006 International Symposium on Distributed Computing and Applications to Business,Engineering and Science(2006年国际电子、工程及科学领域的分布式计算应用学术研讨会)

会议地点: 杭州

会议语种:英文

页码: 767-770

在线出版日期: 2006-10-12（万方平台首次上网日期，不代表论文的发表时间）