摘要
针对实际值序列和预测值序列均为区间数的组合预测问题,把区间数的左右端点作为两个独立的时序数列,通过引入诱导有序加权调和平均(IOWHA)算子,分别建立以区间数左右端点的倒数离差相关系数最大化为准则的IOWHA算子的变权系数最优组合预测模型,再通过引入偏好系数把多目标最优化模型转化为单目标最优化模型,并尝试给出各模型最优解的性质.最后利用数据进行分析,并和其他单项预测方法或组合预测方法相比较,从而证明该方法确实能够提高区间预测的精度.
In view of the combination forcasting problem that the actual value sequence and the predicted value sequence are intervals, This article establishs the optimal combination forecasting model of variable weight coefficient based on induced ordered weighted harmonic averaging （IOWHA） operator, taking the left and right point interval number as two independent time se- ries, and regarding the maximum of the correlation coefficient of the interval number reciprocal deviation around the endpoint as the criterion, and then we turn multi-objective optimization model into a single objective optimization model through the introduction of preference coeffi- cient, and try to give the properties of optimal solutions. Finally, we compare it with other single prediction method or combination forecast method by the literature data, which shows that this method can indeed improve the interval prediction accuracy.
出处
《合肥学院学报（自然科学版）》
2015年第3期17-22,共6页
Journal of Hefei University :Natural Sciences
基金
国家社会科学基金项目(12BTJ008)资助
