摘要
为了提高短生命周期产品在整个生命周期各个阶段的销量预测能力,在前人研究的基础上,基于模糊理论提出了结合多元线性回归和模糊神经网络的组合预测方法模型。该模型综合考虑了生命周期各阶段的特点,首先对产品的生命周期进行了分析和分段,指出单一方法和传统方法的不足之处。在此基础上针对以上问题设计总体预测方案,不仅可以解决前期历史数据缺乏问题,而且可以解决中后期复杂的非线性预测问题,从而使销量预测模型更通用和更精确。实验结果表明了该模型的有效性。
In order to improve the forecasting capability of sales in various stages of entire lifecycle,based on fuzzy set theory,a method which combined multiple linear regression and fuzzy neural network forecasting methods is proposed.First of all,this model analyzed and separated the product lifecycle based on the character of every subsection,and proposed the insufficiency of single and traditional method.Second,our model could solve not only the lack of historical data problems in early commodity market,but also the complex non-linear prediction problems.So that the sales forecasting model is more general and accurate and the validity of the model is verified by experiment.
出处
《计算机工程与设计》
CSCD
北大核心
2010年第11期2527-2530,共4页
Computer Engineering and Design
关键词
生命周期
模糊集理论
线性回归
模糊神经网络
销量预测
lifecycle
fuzzy set theory
linear regression
fuzzy neural network
sales forecasting
