Multilayered feed forward neural network based on particle swarmopti mizer algorithm

• 作 者: 潘峰 陈杰 涂序彦 付继伟
• 机 构: Dept .of Automatic Control ,School of Information Science and Technology ,Beijing Inst .of Technolog
• 出 处: Journal of Systems Engineering and Electronics
• 发表日期: 2005年03期
• 关键词: deviation parti swarm searching otherwise fitness scaled conjugate normalized Fl

BP is a commonly used neural network training method, which has some disadvantages, such as local minima, sensitivity of initial value of weights, total dependence on gradient information. This paper presents some methods to train a neural network, including standard particle swarm optimizer (PSO), guaranteed convergence particle swarm optimizer (GCPSO), an improved PSO algorithm, and GCPSO BP, an algorithm combined GCPSO with BP. The simulation results demonstrate the effectiveness of the three algorithms for neural network training.

BP has a commonly used neural network training method, which has some disadvantages, such as local minima, sensitivity of initial value of weights, total dependence on gradient information. This paper presents some methods to train a neural network, including standard particle swarm optimizer ( PSO), an improved PSO algorithm, and GCPSO BP, an algorithm combined GCPSO with BP. The simulation results demonstrate the effectiveness of the three algorithms for neural network training.