Guseva T.I. APPLYING MACHINE LEARNING TECHNIQUES TO MANAGE INVENTORY IN RETAIL STORES
DOI: https://doi.org/10.15688/ek.jvolsu.2026.1.12
Tatyana I. Guseva
Candidate of Sciences (Economics), Associate Professor, Head of the Department of Economics and Law, South Ural State University, Prosp. Lenina, 76, 454080 Chelyabinsk, Russian Federation, This email address is being protected from spambots. You need JavaScript enabled to view it. , https://orcid.org/0009-0003-4412-0215
Abstract. Based on the analysis of inventory management methods in retail trade, the paper shows the shortcomings of traditional demand planning methods when faced with a real market environment characterized by complex nonlinear dependencies. The author’s hypothesis about the possibility of using an autoregressive model (AR)that includes integration (I) and moving average (MA) in combination with the LSTM machine learning method has been developed and tested in a model for demand forecasting in the retail trade of stationery products by OOO Kantstanta (LLC). The presented model allows processing stationary and non-stationary data series and identifying possible seasonal components and hidden patterns in the data. Sales information is loaded from a CSV file, where the data is presented in a wide format with monthly details, and there is an option for integration with 1C Accounting. The ARIMA model automatically decides whether differentiation is necessary in the case of a nonstationary series. Using LSTM machine learning is an option to improve this model, as machine learning methods show advantages for short-term forecasts, especially under high volatility conditions and in the presence of nonlinear relationships. An important feature of the model is the visual representation of time series with forecast values. The analysis of seven product categories using the model revealed different demand patterns, each of which requires an individual approach to calculating the optimal inventory level. The study results confirmed the effectiveness of using machine learning methods to manage inventory in conditions of dynamic demand changes and high competition.
Key words: inventory management, retail, demand forecasting, autoregressive models, machine learning.
APPLYING MACHINE LEARNING TECHNIQUES TO MANAGE INVENTORY IN RETAIL STORES by Guseva T.I. is licensed under a Creative Commons Attribution 4.0 International License.
 5_Guseva.pdfURL: https://ges.jvolsu.com/index.php/en/component/attachments/download/2362 | 482 Kb | 86 Downloads |