摘要
文章介绍了最小二乘方法和神经网络方法曲线拟合的原理方法,并列出了最小二乘的多项式拟合和既定函数样式拟合以及BP神经网络拟合的Matlab实现过程,利用数学模型和一组试验数据对三种方法的性能进行验证,最后归纳总结了几种方法的性能特点和适用范围。
The fundamental roles of least square curve fitting law and NN curve fitting law are introduced, their realization methods based Matlab are researched. The three ways are validated used mathematic model data and test data. In the end, the characteristic and applied condition of the three curve fitting laws are concluded.
出处
《电脑与信息技术》
2009年第2期34-36,共3页
Computer and Information Technology
