对于季节变动时间序列提出了一种新的乘法分解模型,它可表示成年度变动因子、季节变动因子、总平均值和随机变动因子的乘积.然后在此基础上给出渐进预测模型和乘法预测模型,最后进行了实例分析.
A new multiplicative decomposable model is put forward on seasonal variable time series, which is expressed by the product of four factors that are annual variable factor, seasonal variable factor, general mean value and random variable factor. Then asymptotic forecasting model and multiplicative forecasting model are proposed based on the muhiplicative decomposable expression. Finally, an example is illustrated to show its satisfactory result.