基于数据驱动的电子产品回收量预测研究
发布时间：2018-12-15 21:46
【摘要】：电子产品型号类型多、更新速度快等特点导致了预测回收量的难度大,因此对第三方维修服务商而言,电子产品回收量预测的精度直接影响到企业的运营成本以及服务水平.通过企业真实数据的回归分析,发现产品的累计销售量与累计回收量之间存在显著的线性相关性,由此设计了回归预测方法与阻尼趋势预测方法相结合的组合预测方法,并进行数值实验.实验结果表明该组合预测方法在电子产品回收预测量中能达到比使用单个模型更好的效果,实现了预测精度的显著提升.
[Abstract]:Many types of electronic products and rapid renewal speed lead to the difficulty of predicting the amount of recovery, so for third-party maintenance service providers, the precision of forecasting the amount of recovery of electronic products has a direct impact on the operating costs and service level of enterprises. Through the regression analysis of the real data of the enterprise, it is found that there is a significant linear correlation between the cumulative sales volume and the cumulative recovery amount of the product. Therefore, a combined forecasting method combining regression forecasting method and damping trend forecasting method is designed. Numerical experiments were carried out. The experimental results show that the combined forecasting method can achieve better results than using a single model in the prediction of electronic product recovery, and achieve a significant improvement in prediction accuracy.
【作者单位】： 厦门大学管理学院;厦门大学信息与网络中心;
【基金】：国家自然科学基金(71371158,71671151,71711530046)
【分类号】：F713.2;O212.1
本文编号：2381343
[Abstract]:Many types of electronic products and rapid renewal speed lead to the difficulty of predicting the amount of recovery, so for third-party maintenance service providers, the precision of forecasting the amount of recovery of electronic products has a direct impact on the operating costs and service level of enterprises. Through the regression analysis of the real data of the enterprise, it is found that there is a significant linear correlation between the cumulative sales volume and the cumulative recovery amount of the product. Therefore, a combined forecasting method combining regression forecasting method and damping trend forecasting method is designed. Numerical experiments were carried out. The experimental results show that the combined forecasting method can achieve better results than using a single model in the prediction of electronic product recovery, and achieve a significant improvement in prediction accuracy.
【作者单位】： 厦门大学管理学院;厦门大学信息与网络中心;
【基金】：国家自然科学基金(71371158,71671151,71711530046)
【分类号】：F713.2;O212.1
【相似文献】
相关博士学位论文 前1条
1 付小勇;废旧电子产品回收处理中的博弈模型研究[D];大连理工大学;2012年
，本文编号：2381343
本文链接：https://www.wllwen.com/kejilunwen/yysx/2381343.html
