| 摘要: |
| 针对用区间型数据描述不确定现象的组合预测问题，为了提高区间型数据的预测精度，首先采用诱导有序加权连续区间的广义有序加权平均（IOWC-GOWA）算子将区间数集结为实数；然后对集结后的实数进行标准化处理；最后从信息论的角度引入相对熵作为最优准则，提出了基于IOWC-GOWA算子及相对熵的区间型组合预测模型；另外，通过实例分析了该组合预测模型的合理性和有效性；结果表明：该组合预测模型可以有效地提高区间型数据的预测精度，即该模型是合理有效的，并且，参数λ和BUM函数的选取会对模型的预测精度产生一定的影响。 |
| 关键词: IOWC-GOWA算子 相对熵 标准化 区间型组合预测 |
| DOI： |
| 分类号: |
| 基金项目: |
|
| Interval Combination Forecasting Model Based on IOWC-GOWA Operator and Relative Entropy |
|
DU Kang, YUAN Hong-jun
|
| Abstract: |
| Aiming at combination forecasting problems of describing uncertainty by interval numbers, in order to improve the forecasting accuracy of interval data, firstly, the induced ordered weighted continuous generalized ordered weighted averaging (IOWC-GOWA) operator is used to aggregate the interval numbers into real numbers, then we normalize them. Finally, by introducing relative entropy as the optimum criterion from the perspective of information theory, an interval combination forecasting model based on IOWC-GOWA operator and relative entropy is proposed. Furthermore, the rationality and effectiveness of the combined forecasting model is analyzed by an example. The results show that the combination forecasting model can effectively improve the forecasting accuracy of interval data. That is to say，the model is reasonable and effective. Additionally, the selection of the parameter λand the BUM function has a certain impact on the prediction accuracy of the model. |
| Key words: IOWC-GOWA operator relative entropy normalization interval combination forecasting |
