【摘要】 利用线性神经网络进行协整系数的估计,首先检验变量的数据生成过程,再利用线性神经网络进行拟合,最后检验残差的平稳性,以确定变量之间是否存在协整关系.该方法具有简便、直观、精度高的特点.更多还原

【Abstract】 This paper applies the Linear Neural Network to the estimation of coefficient.It starts with testing of the data generating process of variables,moves on to simulation of the relationship by Linear Neural Network,and ends with testing of the stability of residual to determine the existence of a cointegration relationship between variables.This method is characteristic of being concise,illustrative and highly precise.更多还原