| 基于LM—BP神经网络的GDP预测模型及其应用 |
| |
| 引用本文： | 张自敏,樊艳英,陈冠萍.基于LM—BP神经网络的GDP预测模型及其应用[J].广西梧州师范高等专科学校学报,2014(2):131-135. |
| |
| 作者姓名： | 张自敏 樊艳英 陈冠萍 |
| |
| 作者单位： | [1]贺州学院教育技术中心,广西贺州542899 [2]贺州学院计算机科学与信息工程学院,广西贺州542899 |
| |
| 基金项目： | 贺州学院自然科学研究资助项目（2011zRKY02）、广西高校科学技术研究项目（2013LX143）. |
| |
| 摘 要： | 讨论了基于Levenberg—Marquardt（LM）算法的BP神经网络及其GDP预测的应用。LM算法利用误差函数二阶导数信息，对高斯一牛顿法的优化，相对传统的负梯度法而言，其收敛速度更快。最后以贺州市GDP为例，就预测的效率和精确度来说，LM—BP网络预测GDP的速度和精度明显优于标准的BP算法网络。
|
| 关 键 词： | BP神经网络 LM算法 预测 |
GDP Prediction Model and Its Application Based on LM-BP Neural Network |
| |
| ZHANG Zi-min,FAN Yan-ying,CHEN Guan-ping.GDP Prediction Model and Its Application Based on LM-BP Neural Network[J].Journal of Wuzhou Teachers College of Guangxi,2014(2):131-135. |
| |
| Authors: | ZHANG Zi-min FAN Yan-ying CHEN Guan-ping |
| |
| Institution: | 1.Center of Education Technology, Hezhou University, Hezhou Guangxi 542899; 2.School of Computer Science and Information Engineering, Hezhou University, Hezhou Guangxi 542899) |
| |
| Abstract: | BP neural network based on Levenberg-Marquardt （LM） algorithm and its application on predicting GDP is discussed in this paper. With the error function of second derivative information, the LM-BP algorithm network is faster than the gradient method. Taking Hezhou city GDP as an example, the LM-BP network is superior to the standard BP network, which is concerned with efficiency and accuracy of prediction. |
| |
| Keywords: | BP neural network LM algorithm prediction |
| 本文献已被 维普 等数据库收录！ |
