摘要
将人工神经网络引入回归分析过程，探讨了回归分析神经网络的结构和学习算法，研究了基于人工神经网络的模型变量的选择、观测样本的采集和使用等．进行了仿真实验，仿真结果初步显示了神经网络方法能够较好地解决传统的回归方法所面临的困难，并具有较高的模型精度．
By introducing artificial neural network to regression analysis processes,this article discusses the construction of regression analysis neural network and learningalgorithm, studies the choice of variables in the models based on artificial neuralnetworks, and the collection and use of observed samples. An experiment on emulationis carried out which proves that preliminarily the neural networks method can solve theproblems better than the traditional regression method; moreover a higher degree of precision is achieved.
出处
《北方工业大学学报》
1999年第1期68-73,共6页
Journal of North China University of Technology
关键词
神经网络
非线性回归
预测模型
参数估计
artificial neural networks
non-linear regression
forecast model
