Dynamic Pricing and Replenishment Strategy for Supermarket Vegetables Based on LSTM and Bi-Objective Optimization

Fangyan Ma

GBP Proceedings Series, 2025, vol. 15, 1-9

Abstract: To address the dual operational objectives of "profit maximization" and "loss minimization" in supermarket vegetable management, this study proposes a three-stage decision-making framework integrating LSTM-based sales forecasting, price elasticity correction, and an Improved Particle Swarm Optimization (IPSO)-driven bi-objective optimization model. The framework was validated using 180 days of operational data (January 1-June 30, 2024) from six representative vegetable categories-lettuce, tomato, cucumber, carrot, potato, and spinach-collected across three supermarkets of different scales: Supermarket A (RT-Mart, a regional leading chain), Supermarket B (Liyoumart, a community-based chain), and Supermarket C (a local convenience store). Data were obtained from the supermarkets' ERP systems and on-site inventory records. Experimental results reveal that: (1) the LSTM-based sales forecasting model reduces the Mean Absolute Error (MAE) by 23.6% compared with the traditional ARIMA model, achieving a coefficient of determination (R 2 ) of 0.92, thereby demonstrating strong capability in capturing nonlinear fluctuations in sales. (2) Significant differences in price elasticity are observed across both vegetable categories and supermarket types: leafy vegetables (lettuce and spinach) show elastic demand (|ε| = 1.82-2.15), while root vegetables (carrot and potato) exhibit inelastic demand (|ε| = 0.78-0.95). Moreover, smaller supermarkets (e.g., Supermarket C) display 8%-10% higher absolute elasticity values than larger ones (e.g., Supermarket A). (3) For key predictive tasks, the average MAE of replenishment quantity prediction is 8.2 kg, while the Mean Absolute Percentage Error (MAPE) for price prediction is 4.5%, both meeting the operational benchmark of "prediction error ≤ 10%." (4) The optimized strategy improves the average comprehensive profit of vegetables across the three supermarkets by 16.8% (18.7% for A, 17.2% for B, and 14.5% for C) and reduces the average loss rate from the industry-typical 12.3% to 6.8%, with the most significant reduction observed in leafy vegetables (from 15.2% to 5.9%). This study provides a quantifiable and practical decision-support framework for intelligent vegetable supply chain management applicable to supermarkets of varying scales.

Keywords: supermarket vegetables; LSTM neural network; price elasticity; bi-objective optimization; replenishment-pricing prediction (search for similar items in EconPapers)
Date: 2025
References: Add references at CitEc
Citations:

Downloads: (external link)
https://soapubs.com/index.php/GBPPS/article/view/866/844 (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:axf:gbppsa:v:15:y:2025:i::p:1-9

Access Statistics for this article

More articles in GBP Proceedings Series from Scientific Open Access Publishing
Bibliographic data for series maintained by Yuchi Liu ().