基于异方差的递增时间序列自回归预测的改善方法

Improvement of autoregressive prediction method based on increasing time-series data of heteroscedasticity

依据"利用反双曲正弦函数变换提高数据列光滑程度"的结论,给出了两种异方差条件下递增时间序列自回归预测的改善方法.

According to the conclusion that the smoothness level of a data row can be increased through the transform of the counter-hyperbolic sine function, two methods of autoregression forecasting based on the increasing time series are given.