一种简易预测模型的应用

Application of A Simple Forecasting Model

预测是各个部门制订工作计划的一项基本工作 ,是指对未来活动的科学推测。预测的正确与否取决于对预测模型的分析。而预测的方法有“最小二乘法”、“卡尔曼滤波法”、“最小方差预测法”、“指数平滑法”等等 ,但是以上方法的特点是模型复杂 ,计算工作量大。据此介绍了两种比较简单的预测模型即“灰色预测模型”及“递推及残差识别预测模型” 。

Forecast is a basic procedure in working out a plan, is a scientific prediction for the activities in the future. The analysis on forecasting model is very critical for the correctness of the forecast. There is several methods could be used for forecasting, such as least square procedure, Kalman filter model, minimum variance forecasting and index smoothing forecasting etc. These methods share the characteristics of complex model and big amount of calculation. Two relevant simple methods namely gray prediction, recurrence and residual error identification forecasting model were introduced. Moreover, the forecasting results achieved by these two methods were the same.