With the increasing number of merchants on e-commerce platforms and the continuous enhancement of goods quantity, inventory warehouses in various regions are facing growing warehousing pressure. Therefore, e-commerce platforms need to reduce the warehouse inventory pressure through scientific and effective management means and optimizing the supply chain based on big data application scenarios. This paper analyzes the demand forecasting and inventory optimization issues of e-commerce retail merchants based on the SARIMA model and fuzzy comprehensive evaluation. For the problem, this article first merges the data and organizes the time series data of the same merchant, same warehouse, and same product. Subsequently, the data is processed to remove two outlier values. Considering that the demand for each product is affected by seasonal components, this paper uses the SARIMA model to forecast the time series data and saves the forecast results in the result table. Next, the paper evaluates the forecasting performance of the SARIMA model using fuzzy comprehensive evaluation and calculates the comprehensive evaluation value for the time series data of the same merchant, the same warehouse, and the same product. Based on the calculated comprehensive evaluation value and the 1-wmape index, it concludes that the SARIMA model has good forecasting performance for Problem 1's time series data and the forecasting results are relatively accurate. Finally, this paper classifies the time series with similar comprehensive evaluation values into one category, making the demand characteristics of the same category as similar as possible, and ultimately divides the time series into products with high demand and products with low demand.In conclusion, we evaluated the advantages and disadvantages of the models and methods used in this paper, hoping that future research can further improve these models, enhancing their performance and applicability. Similarly, similar problems can refer to the methods and ideas of this paper for analogous research and analysis.

Jinyu Liu, Zhewei Zhang, Qiming Song, Binyang Luo. Study on Demand Forecasting and Inventory Optimization Using SARIMA and Fuzzy Analysis. The Frontiers of Society, Science and Technology (2023) Vol. 5, Issue 18: 104-110. https://doi.org/10.25236/FSST.2023.051818.

[1] Huang Yi, Yang Yongsheng, Zhang Qingguang. Application of the SARIMA Model in Monthly Average Temperature Time Series [C]// Strengthening Scientific and Technological Foundations to Promote Meteorological Modernization—The 29th Annual Meeting of the Chinese Meteorological Society. 2012.

[2] Han Zhonggeng. Mathematical Modeling Methods and Applications [M]. Higher Education Press, 2017. 

[3] Zhang Li, Niu Huifang. Forecast Analysis of Consumer Price Index Based on SARIMA Model [J]. Journal of Applied Statistics and Management, 2013, 32(1):6. DOI:CNKI:SUN:SLTJ.0.2013-01-004.

[4] Yang Jinfang, Zhai Yongjie, Wang Dongfeng, et al. Time Series Prediction Based on Support Vector Regression [J]. Proceedings of the CSEE, 2005, 25(17):5. DOI:10.3321/j.issn: 0258-8013. 2005. 17.022.