[1]张志伟,胡伍生,黄晓明.线性回归模型精化方法[J].东南大学学报(自然科学版),2009,39(6):1279-1282.[doi:10.3969/j.issn.1001-0505.2009.06.037]
 Zhang Zhiwei,Hu Wusheng,Huang Xiaoming.Linear regressive model improved by neural network[J].Journal of Southeast University (Natural Science Edition),2009,39(6):1279-1282.[doi:10.3969/j.issn.1001-0505.2009.06.037]
点击复制

线性回归模型精化方法()
分享到:

《东南大学学报(自然科学版)》[ISSN:1001-0505/CN:32-1178/N]

卷:
39
期数:
2009年第6期
页码:
1279-1282
栏目:
其他
出版日期:
2009-11-20

文章信息/Info

Title:
Linear regressive model improved by neural network
作者:
张志伟 胡伍生 黄晓明
东南大学交通学院, 南京 210096
Author(s):
Zhang Zhiwei Hu Wusheng Huang Xiaoming
School of Transportation, Southeast University, Nanjing 210096, China
关键词:
模型精化 趋势项 回归残差 神经网络
Keywords:
improving model tendency item regressive residual neural network
分类号:
P413
DOI:
10.3969/j.issn.1001-0505.2009.06.037
摘要:
为了解决由试验观测数据建立的回归拟合模型存在的模型误差,用基于回归残差的神经网络方法精化模型.采用给定方程获得模拟数据,通过数据结构散点图建立回归模型趋势项,利用经典最小二乘法估计趋势项参数,由趋势项参数计算回归残差,借助误差分级迭代的改进BP算法对趋势项进行精化,将两部分叠加获得精化模型.试验结果验证了基于回归残差的神经网络方法精化模型的有效性:神经网络方法精化后的模型能提高回归模型的拟合及预测精度5倍以上,优于最小二乘配置法和半参数法精化结果.神经网络方法精化模型既克服了单一神经网络模型的不可解释性,使模型具有物理意义,又具有较高的预测精度.
Abstract:
The regression fitting model established with the experiment observation data inevitably has the model error. Thus the neural network method based on the regression residual is adopted to improve the model. The simulation data are obtained by the given equation, and the tendency item of regression model is established by the chart of scatter data structure. The tendency parameter is estimated by the classical least squares method, and the regression residual is computed through the tendency parameter. The error grade iterative method of BP(back propagation)neural networks carries on the compensation to the tendency item, and the improved model is obtained by the splicing of the two parts. The results verify the validity of the model improved by the neural network based on the regression residual. The model improved by neural network can improve the regression model fitting, and can improve the forecast accuracy by more than 5 times. It is superior to least squares collocation method and semi-parametric method. The model improved by neural network overcomes the non-explanation in single neural network model, so that the model has physical meaning, and has higher prediction accuracy.

参考文献/References:

[1] 王松桂,史建红,尹素菊,等.线性模型引论[M].北京:科学出版社,2007:1-150.
[2] 陶本藻,施闯,姚宜斌.顾及系统误差的平差模型研究[J].测绘学院学报,2002,19(2):79-81.
  Tao Benzao,Shi Chuang,Yao Yibin.Research on adjustment model with systematical error[J].Journal of Institute of Surveying and Mapping,2002,19(2):79-81.(in Chinese)
[3] 陶本藻.测量数据处理的统计理论和方法[M].北京:测绘出版社,2007:134-192.
[4] 武汉大学测绘学院测量平差学科组.误差理论与测量平差基础[M].武汉:武汉大学出版社,2003:213-223.
[5] 陈希孺,王松桂.近代回归分析[M].合肥:安徽教育出版社,1987:217-249.
[6] 胡伍生.神经网络理论及其应用[M].北京:测绘出版社,2006:63-113.
[7] 黄维彬.近代测量平差理论及其应用[M].北京:解放军出版社,1992:105-189.
[8] 胡伍生,沙月进.神经网络BP算法的误差分级迭代法[J].东南大学学报:自然科学版,2003,33(3):376-378.
  Hu Wusheng,Sha Yuejin.Error grade iterative method of BP neural networks [J]. Journal of Southeast University:Natural Science Edition,2003,33(3):376-378.(in Chinese)
[9] 丁士俊.测量数据的建模与半参数估计[D].武汉:武汉大学测绘学院,2005.
[10] 丁士俊,陶本藻.自然样条半参数模型与系统误差估计[J].武汉大学学报:信息科学版,2004,29(11):964-967.
  Ding Shijun,Tao Benzao.Semiparametric regression model with natural spline and systematic error estimation[J].Geomatics and Information Science of Wuhan University,2004,29(11):964-967.(in Chinese)

备注/Memo

备注/Memo:
作者简介: 张志伟(1964—),男, 博士生,副教授; 黄晓明(联系人),男,博士,教授,博士生导师,huangxm@seu.edu.cn.
基金项目: 国家高技术研究发展计划(863计划)资助项目(2007AA12Z228).
引文格式: 张志伟,胡伍生,黄晓明.线性回归模型精化方法[J].东南大学学报:自然科学版,2009,39(6):1279-1282. [doi:10.3969/j.issn.1001-0505.2009.06.037]
更新日期/Last Update: 2009-11-20