基于IOWA算子的多参数指数平滑模型组合预测

Integrated Prediction of Multi Parameter Exponential Smoothing Model Based on IOWA Operator

为了提高预测精度,文章引入了诱导有序加权平均算子,首先对观测值进行单项预测,然后分别构建了以误差平方和最小、绝对误差绝对值之和最小为准则的IOWA组合预测模型,通过求解数学规划模型给出了两准则下的IOWA组合预测最优权系数。最后,以中关村高新技术园区为例进行实证分析,结果表明了模型的有效性。

In order to improve the prediction accuracy, this paper introduces the induced ordered weighted average operator. Firstly the paper makes a single prediction of the observed values, and then respectively constructs the IOWA combined forecasting model with the minimum error sum of squares and minimum sum of absolute error absolute value as the criterion. The paper also gives the optimal weight coefficient of IOWA combined forecasting under the two criteria by solving mathematical programming model. Finally, the paper takes Zhongguancun High-tech Park as an example to make an empirical analysis, and the result shows the validity of the model.