| 基于改进PSO算法的BP神经网络的应用研究 |
| |
| 引用本文： | 曾万里 危韧勇 陈红玲. 基于改进PSO算法的BP神经网络的应用研究[J]. 微机发展, 2008, 18(4): 49-51 |
| |
| 作者姓名： | 曾万里 危韧勇 陈红玲 |
| |
| 作者单位： | 中南大学信息科学与工程学院 湖南长沙410075(曾万里，危韧勇)，吉林大学珠海学院 广东珠海519041(陈红玲) |
| |
| 摘 要： | 为了克服粒子群优化算法本身存在的早熟和局部收敛的固有问题，在描述了BP神经网络的基本结构的基础上，介绍了粒子群优化算法（PS0）的基本概念，并通过对二者优缺点的分析与比较，结合二者的优势，将粒子矢量位移应用到PS0算法中，并在此基础上，用改进的PS0算法对BP网络进行训练，还利用某商场的部分消费数据进行了实验。结果表明，基于改进的PS0算法的BP网络在收敛速度和精度上都比基于传统的PSO算法好。
|
| 关 键 词： | Bp网络 粒子群优化算法 矢量位移 收敛速度 |
| 文章编号： | 1673-629X(2008)04-0049-03 |
| 修稿时间： | 2007-07-02 |
Research and Application of BP Neural Network Based on Improved PSO Algorithm |
| |
| ZENG Wan-li,WEI Ren-yong,CHEN Hong-ling. Research and Application of BP Neural Network Based on Improved PSO Algorithm[J]. Microcomputer Development, 2008, 18(4): 49-51 |
| |
| Authors: | ZENG Wan-li WEI Ren-yong CHEN Hong-ling |
| |
| Affiliation: | ZENG Wan-li1,WEI Ren-yong1,CHEN Hong-ling2 |
| |
| Abstract: | Describes the basic structure of BP neural network in order to solve the inhere problem of precocity and refraining partly.Some basic concepts of particle swarm optimizer are introduced also.Combines the advantages of them through comparing and analyzing some problems concerned to them.What's more,the concept of particle's vector shift is used in algorithm PSO.Based on this,improved algorithm PSO is used to train BP neural network.This experiment is done with a market's consumption data.As a result,it is better to use improved algorithm PSO than traditional algorithm PSO or BP alone not only in velocity but in precision of convergence. |
| |
| Keywords: | BP neural network particle swarm optimizer vector shift velocity of convergence |
| 本文献已被 CNKI 维普 等数据库收录！ |
