| 改进粒子群算法对BP神经网络的优化 |
| |
| 引用本文： | 沈学利,张红岩,张纪锁. 改进粒子群算法对BP神经网络的优化[J]. 计算机系统应用, 2010, 19(2): 57-61 |
| |
| 作者姓名： | 沈学利 张红岩 张纪锁 |
| |
| 作者单位： | 辽宁工程技术大学,电子与信息工程学院,辽宁,葫芦岛,125105 |
| |
| 摘 要： | 介绍一种基于改进粒子群算法优化BP网络的权值调整综合方法。该算法在传统BP算法的误差反传调整权值的基础上，引入粒子群算法的权值修正，并且在训练神经网络权值的同时优化其连接结构，删除冗余连接，从而建立了基于粒子群算法优化的BP网络新模型。结果表明，改进算法不仅可以克服传统BP算法收敛速度慢和易陷入局部权值的局限，而且很大程度地提高了结果精度和BP网络学习能力。
|
| 关 键 词： | 粒子群算法 惯性权值 神经网络 BP算法 优化 |
| 收稿时间： | 2009-06-08 |
Improved Particle Swarm Optimization Algorithm |
| |
| SHEN Xue-Li,ZHANG Hong-Yan and ZHANG Ji-Suo. Improved Particle Swarm Optimization Algorithm[J]. Computer Systems& Applications, 2010, 19(2): 57-61 |
| |
| Authors: | SHEN Xue-Li ZHANG Hong-Yan ZHANG Ji-Suo |
| |
| Affiliation: | SHEN Xue-Li,ZHANG Hong-Yan,ZHANG Ji-Suo(School of Electronic , Information Engineering,Liaoning Technical University,Huludao 125105,China) |
| |
| Abstract: | A new method to adjust weights of BP network is proposed.The new model is based on the weight adjustments of traditional BP algorithm by tuning the structure and connection weights of BP network and improved particle swarm optimization simultaneously.The result shows that the improved algorithm can not only overcome the limitations in both the slow convergence and the local extreme values of traditional BP algorithm,but also improve the precision of the result and the learning ability greatly. |
| |
| Keywords: | particle swarm optimization inertia weight neural networks back propagation arithmetic optimization |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
| 点击此处可从《计算机系统应用》浏览原始摘要信息 |
|
点击此处可从《计算机系统应用》下载全文 |
