中国计量大学学报2017，Vol.28Issue(3)：334-339,6.DOI:10.3969/j.issn.2096-2835.2017.03.011
二次指数平滑法优化马尔科夫预测模型
Optimization on Markov prediction models with the second exponential smoothing method
摘要
Abstract
Markov has "no aftereffect" ,which means that future sales will only be related to the current sales data ,not to the past sales figures .In fact ,different time points in the past have different degrees of impact on current sales results . The exponential smoothing , which makes up for the shortcomings of the Markov prediction model ,suggests that recent past sales data ,in some ways ,will influence the future .In this paper , the quadratic exponential smoothing coefficient method is used to optimize the Markov prediction model ,and is verified with some brand electric car sales .It was found that the absolute error of the optimized prediction model was less than the predicted results of Markov models .Therefore ,it is concluded that the Markov prediction model based on the quadratic exponential smoothing method is feasible .关键词
马尔科夫链/状态转移概率/二次指数平滑法/销售预测Key words
Markov Chain/transition probability matrix/secondary exponential smoothing/sales forecast分类
信息技术与安全科学引用本文复制引用
吕丹丹,顾巧祥,邢超..二次指数平滑法优化马尔科夫预测模型[J].中国计量大学学报,2017,28(3):334-339,6.基金项目
浙江省自然科学基金资助项目(No.LY14E050024). （No.LY14E050024）
