摘要
企业可以持续不断地收集市场需求信息,利用贝叶斯方法获得更精确的市场需求预测。通常订货时间越晚,收集的市场需求信息数据越多且越精确,然而此时采购成本会较高。企业需要决策在当前时点是否订货以及最优的订货量。文章运用搜集的数据建立市场需求的预测过程,分析数据对预测市场需求的影响。当市场需求预测的调整量为正态分布且存在订货成本时,建立了基于贝叶斯预测更新的报童模型,并给出了企业最优库存策略和求解算法。通过敏感性分析说明了采购成本和市场需求信息对库存策略的影响。
Enterprises can continuously collect market demand information and use Bayesian method to acquire more accu- rate market demand prediction. Usually the later the order is made, the richer and more precise the collected market demand infor- mation data is. But at this time the purchasing cost is relatively higher. Thus enterprises have to determine whether or not to order and order the optimum quantity as well at the current time-point. This paper employs the collected data to establish market de- mand forecast process and analyze the impact of data on market demand forecast. When the adjustment of market demand forecast is of a normal distribution and there exists ordering cost, the paper establishes a newsvendor model based on Bayesian forecast evolution, and gives an optimal inventory strategy and the solution algorithm. Finally the paper gives a sensitive analysis to illus- trate the impact of purchasing cost and market demand information on inventory strategy.
出处
《统计与决策》
CSSCI
北大核心
2017年第24期39-43,共5页
Statistics & Decision
基金
国家自然科学基金资助项目(71271168
71472140)
教育部人文社会科学基金资助项目(10YJC630028)
关键词
预测更新过程
市场需求
贝叶斯
报童模型
forecast evolution process
market demand
Bayesian
newsvendor model
