基于Markov模型的中国高端卷烟产品销量预测研究

本文通过对全国调拨价每条100元以上的卷烟销量进行分析,预测出未来几年内其销量增长的趋势。笔者依据历年销量的增长率不同将其划分为不同的状态,计算出状态之间的转移概率矩阵,并构造出销量增长率的马尔可夫模型,求出其销量增长率的期望值,通过对比近两年的实际销量值计算出其预测销量的误差,为未来几年卷烟销量的预测提供参考,同时此模型适用于中国中低端卷烟的销量预测,研究结果对于全国烟草行业的营销实践具有参考价值。

Research on Ultra Cigarette Sales Prediction Based on Markov Model in China

In order to predict the tendency of tobacco sale growth in following years,the article uses the national tobacco sale's data to analyses,which the tobacco transfer price tagged with $100 or up $100.Thus,we firstly divided the previous sale's growth rate into different conditions and calculated the transition probability matrix between the different states.Secondly,we structured the Markov model of the sale's growth rate and gave out their expectations.Finally we figured out the predicted sale's errors in order to provide the references to the sale's forecast in following years,which is through contrasting the actual sale's vale in recent two years.At the same time,the Markov model can also be applied to predict the low-end tobacco sale's tendency and the result will have reference value to national tobacco industry marketing.