应用马尔可夫链模型预测零售业库存需求的精准度研究

Research on the Accuracy of Applying Markov Chain Model to Forecast Retail Industry Inventory Demand

摘要: 随着零售业的发展，库存管理成为零售业运营中非常重要的一环，一定程度上影响到企业资金链，进而影响运营效率。因此精准预测库存需求量对于实现零售新物流改革十分重要。本文提出利用马尔可夫链模型来预测库存数据，通过选取M超市14个品类在2021年12个月的库存量数据进行分析研究。由于各品类量纲不同，首先将库存数据进行Z-Score标准化，结合历史数据采用马尔可夫过程计算每月库存状态转移概率，并对下月库存状态进行预测，将预测数据与当月实际数据进行对比。结果表明：应用马尔可夫链模型预测的库存数据准确度高达85.71%，该模型在一定程度上可以帮助管理者更好地进行库存管理，减少企业存货成本管理中的缺货成本，库存成本等主要问题，提高物流平台的运行质量。

Abstract: With the development of the retail industry, inventory management has become a very important part of the operation of the retail industry, affecting the enterprise’s capital chain to a certain ex-tent, and thus affecting the operation efficiency. Therefore, accurate prediction of inventory demand is very important for realizing the new retail logistics reform. This paper proposes to use the Markov chain model to predict the inventory data, and analyzes the inventory data of 14 categories of M Supermarket in the 12 months of 2021. Due to the different dimensions of each category, the inventory data is first standardized by Z-Score, and the Markov process is used to calculate the transfer probability of monthly inventory status in combination with historical data, and the inventory status of the next month is predicted. The predicted data is compared with the actual data of the current month. The results show that the accuracy of the inventory data predicted by the Markov chain model is up to 85.71%. The model can help managers to better manage inventory, reduce the main problems such as shortage cost and inventory cost in enterprise inventory cost management, and improve the operation quality of the logistics platform.