Author Login Chief Editor Login Reviewer Login Editor Login Remote Office

›› 2004, Vol. 30 ›› Issue (01): 54-.

• Papers • Previous Articles     Next Articles

Simplified Generalized Multi-layer Perceptron Model and Its Learning Algorithm

FANG Ning,LI Jingzhi,HE Guiming   

  1. (State Key Laboratory of Software Engineering, Wuhan University, Wuhan 430072 )
  • Published:2004-01-15

简化的广义多层感知机模型及其学习算法

方宁,李景治,贺贵明   

  1. 武汉大学软件工程国家重点实验室,武汉大学软件工程国家重点实验室,武汉大学软件工程国家重点实验室 武汉 430072

Abstract: This paper sets up a simplified generalized multi-layer perceptron model (SGMLP model). Two learning algorithms are proposed to train SGMLP network s. One is the generalized error back propagation algorithm (GBP), and the other is based on genetic algorithm (GA). Finally, experimental results of two typical benchmarks demonstrate that the new model and its learning algorithm are feasib le and efficient.

摘要: 提出了简化的广义多层感知机模型(SGMLP模型),并针对SGMLP模型给出了两种 学习算法:广义误差反向传播算法(GBP算法)和基于遗传算法(GA)的学习算法。两个典 型算例的实验结果表明,该模型及其学习算法是可行和有效的。

关键词: 简化的广义多层感知机, 遗传算法, 广义误差反向传播算法

CLC Number: