| 基于复合适应度微粒群算法的神经网络训练 |
| |
| 引用本文： | 赵 辉,刘鲁源,张更新. 基于复合适应度微粒群算法的神经网络训练[J]. 控制与决策, 2005, 20(8): 958-960 |
| |
| 作者姓名： | 赵 辉 刘鲁源 张更新 |
| |
| 作者单位： | 天津大学,电气与自动化工程学院,天津,300072;天津理工大学,自动化学院,天津,300191;天津大学,电气与自动化工程学院,天津,300072;天津理工大学,自动化学院,天津,300191 |
| |
| 基金项目： | 天津市自然科学基金重点项目(033803311)，天津市高等学校科技发展基金项目(20041705) |
| |
| 摘 要： | 为提高神经网络的泛化能力,针对以均方误差为适应度的PSO算法在训练神经网络时会产生一定的过拟合问题,提出对均方误差和误差分布均匀度进行信息融合,构成复合适应度作为训练指标.实验结果表明,该方法可使网络的泛化能力得到明显的改善.
|
| 关 键 词： | 微粒群算法 神经网络 复合适应度 泛化能力 |
| 文章编号： | 1001-0920(2005)08-0958-03 |
| 收稿时间： | 2004-09-13 |
| 修稿时间： | 2004-09-13 |
Neural Network Training Based on Compound Fitness Particle Swarm Optimization |
| |
| ZHAO Hui,LIU Lu-yuan,ZHANG Geng-xin. Neural Network Training Based on Compound Fitness Particle Swarm Optimization[J]. Control and Decision, 2005, 20(8): 958-960 |
| |
| Authors: | ZHAO Hui LIU Lu-yuan ZHANG Geng-xin |
| |
| Abstract: | The over fitting arises if the PSO(particle swarm optimization) algorithm whose fitness is mean square deviation is applied in training neural network. In order to improve generalization capacity of feedforward neural network. The compound fitness based on information merging of mean square deviation and error uniformity is proposed as the training index of PSO. The results show that the approach can improve the generalization capacity of feedforward neural networks remarkably. |
| |
| Keywords: | Particle swarm optimization Neural network Compound fitness Generalization ability |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
| 点击此处可从《控制与决策》浏览原始摘要信息 |
|
点击此处可从《控制与决策》下载全文 |
|
