| 注册
首页|期刊导航|哈尔滨商业大学学报(自然科学版)|基于一类损失函数的组合预测模型构建

基于一类损失函数的组合预测模型构建

殷和俊 杨桂元

哈尔滨商业大学学报(自然科学版)2018,Vol.34Issue(1):107-112,6.
哈尔滨商业大学学报(自然科学版)2018,Vol.34Issue(1):107-112,6.

基于一类损失函数的组合预测模型构建

Study on construction of combination forecast model based on loss function

殷和俊 1杨桂元1

作者信息

  • 1. 安徽财经大学统计与应用数学学院,安徽蚌埠233030
  • 折叠

摘要

Abstract

The key problem of combination prediction is to determine the power coefficient of each single item prediction.By constructing a class of loss functions,the minimum loss func-tion as the target is the way to calculate the weight coefficient of each single prediction in the combination forecast model in this paper.And the forecast effects of the corresponding com-bination forecast when the parameters in loss function take different values were compared. In the variable weight combination forecast,it was not only a set of weights can make predic-tion effect to achieve the optimal.

关键词

组合预测/损失函数/预测精度/权系数/模型

Key words

combination forecast/loss function/prediction accuracy/weight coefficient/model

分类

管理科学

引用本文复制引用

殷和俊,杨桂元..基于一类损失函数的组合预测模型构建[J].哈尔滨商业大学学报(自然科学版),2018,34(1):107-112,6.

哈尔滨商业大学学报(自然科学版)

1672-0946

访问量0
|
下载量0
段落导航相关论文