摘要
本文通过比较不同的销售预测方法,指出指数平滑法进行销售预测的优点,并且说明指数平滑法中平滑系数的大小对预测值有较大影响,重点探讨了在一次指数平滑法中如何采用改进的试算法对平滑系数α进行优化,提高预测的准确性。
This paper, through the comparison of different sales forecasting methods, discusses the advantages of using the index level and smooth method to predict the sales, points out the fact that the exponential smoothing coefficient has great effects on the predicted value and expounds the ways to optimize the smoothing coefficient with the improved test algorithm in the single exponential smoothing method to improve the accuracy of prediction.
出处
《宁波工程学院学报》
2011年第4期128-132,共5页
Journal of Ningbo University of Technology
关键词
平滑指数法
平滑系数
试算法
改进
index level and smooth method
smoothing coefficient
test algorithm
improvement
