论文部分内容阅读
本文针对季节性时间序列,提出了自适应季节预测模型。它是在常规季节模型的基础上,对某些系数用自适应指数平滑法进行修正的结果。在银行业务中的运用表明,用这种模型进行预测,按 MAPE 标准,其准确度比常规模型有一个等级的提高。
In this paper, we propose an adaptive seasonal prediction model for seasonal time series. It is based on the conventional seasonal model, the result of some coefficients modified by adaptive exponential smoothing. The use of banking models shows that using this model for forecasting improves the accuracy of the MAPE standard by one level over the conventional model.