基于机器学习的商超收益优化与定价策略研究
Research on Revenue Optimisation and Pricing Strategy of Superstores Based on Machine Learning
摘要: 在生鲜商超中，蔬菜类商品的保鲜期都比较短，且品相随销售时间的增加而变差，隔日就无法再售。本文对多种单品编码进行分析。基于XGBoost模型，通过PSO算法构建以最大商超收益为目标的模型。经过计算，第一种单品编码的最优销售单价为6.299999948531562，最优的销售数量为41.19558，最大商超收益为110.80144372618932；第二种单品编码的最优销售单价为102900011030059，最优的销售数量为30.746147，最大商超收益为74.18823167795038。本研究的一个关键创新点是通过XGBoost模型和PSO算法，实现了对不同蔬菜类商品的个性化最优定价。这意味着商超可以根据每种商品的特性和市场需求，制定最佳的销售单价，以最大化商超的收益。通过更合理的定价和销售策略，商超可以减少蔬菜类商品的滞销和浪费，有助于减少资源浪费，提高经济效益，同时对环保也有积极影响。本文的方法为商超提供了科学的管理决策支持，使其能够更好地应对市场变化和商品特性，提高了经营效益。
Abstract: In fresh food superstores, the freshness period of all vegetable items is relatively short, and the quality deteriorates with the increase of selling time, and they cannot be re-sold on the next day. This paper analyses a variety of single-item codes. Based on the XGBoost model, the PSO algorithm is used to construct a model with the goal of maximising supermarket revenue. After calculation, the optimal unit price of the first single product code is 6.29999999948531562, the optimal num-ber of sales is 41.19558, and the maximal hyper-merchandising revenue is 110.80144372618932; the optimal unit price of the second single product code is 102900011030059, the optimal number of sales is 30.746147, and the maximal hyper-merchandising revenue is 74.1818932; the optimal unit price of the second single product code is 102900011030059 and the optimal number of sales is 30.746147. The maximum hypermarket gain is 74.18823167795038. A key innovation of this research is to realize the personalized optimal pricing of different vegetable products through XGBoost model and PSO algorithm. This means that the supermarket can set the best selling unit price according to the characteristics and market demand of each commodity to maximize the profit of the supermarket. Through more reasonable pricing and sales strategies, supermarkets can re-duce the unsalable and waste of vegetable commodities, help reduce resource waste, improve eco-nomic efficiency, and have a positive impact on environmental protection. The method in this paper provides scientific management decision support for the supermarket, so that it can better cope with the market changes and commodity characteristics, and improve the operating efficiency.
文章引用：王哲, 杨渠钏, 卢灏, 陈静琳, 梁兰青, 吴延科. 基于机器学习的商超收益优化与定价策略研究[J]. 计算机科学与应用, 2023, 13(12): 2623-2628. https://doi.org/10.12677/CSA.2023.1312261

参考文献

[1] Mišić, V.V. and Perakis, G. (2020) Data Analytics in Operations Management: A Review. Manufacturing & Service Operations Management, 22, 158-169. [Google Scholar] [CrossRef] 
[2] Chen, X., Owen, Z., Pixton, C. and Simchi-Levi, D. (2022) A Statistical Learning Approach to Personalization in Revenue Management. Management Science, 68, 1923-1937. [Google Scholar] [CrossRef] 
[3] Elmachtoub, A.N., Gupta, V. and Hamilton, M.L. (2021) The Value of Personalized Pricing. Management Science, 67, 6055-6070. [Google Scholar] [CrossRef] 
[4] Qu, T., Zhang, J.H., Chan, F.T., Srivastava, R.S., Tiwari, M.K. and Park, W.Y. (2017) Demand Prediction and Price Optimization for Semi-Luxury Supermarket Segment. Computers & Industrial Engineering, 113, 91-102. [Google Scholar] [CrossRef] 
[5] Guo, S., Chen, C., Wang, J., Liu, Y., Xu, K., Yu, Z., et al. (2019) Rod-Revenue: Seeking Strategies Analysis and Revenue Prediction in Ride-on-Demand Service Using Multi-Source Ur-ban Data. IEEE Transactions on Mobile Computing, 19, 2202-2220. [Google Scholar] [CrossRef] 
[6] Greenstein-Messica, A. and Rokach, L. (2020) Machine Learn-ing and Operation Research Based Method for Promotion Optimization of Products with No Price Elasticity History. Electronic Commerce Research and Applications, 40, Article ID: 100914. [Google Scholar] [CrossRef] 
[7] Chen, S.S., Choubey, B. and Singh, V. (2021) A Neural Net-work Based Price Sensitive Recommender Model to Predict Customer Choices Based on Price Effect. Journal of Retail-ing and Consumer Services, 61, Article ID: 102573. [Google Scholar] [CrossRef] 
[8] Pokhariya, J., Mishra, P.K. and Kandpal, J. (2022) Machine Learning for Intelligent Analytics. CRC Press, Boca Raton, 9-234. [Google Scholar] [CrossRef] 
[9] Keskin, N.., Li, Y. and Song, J.S. (2022) Data-Driven Dynamic Pricing and Ordering with Perishable Inventory in a Changing Environment. Management Science, 68, 1938-1958. [Google Scholar] [CrossRef] 
[10] Shum, S., Tong, S. and Xiao, T. (2017) On the Impact of Uncertain Cost Reduction When Selling to Strategic Customers. Management Science, 63, 843-860. [Google Scholar] [CrossRef] 
[11] Zhao, J. (2021) Synergy between Customer Segmentation and Per-sonalization. Journal of Systems Science and Systems Engineering, 30, 276-287. [Google Scholar] [CrossRef] 
[12] Wei, W., Feng, L. and Liu, C. (2018) Mixed Causal Structure Discovery with Application to Prescriptive Pricing. Proceedings of the 27th International Joint Conference on Artificial Intelligence, Stockholm 13-19 July 2018, 5126-5134. [Google Scholar] [CrossRef] 
[13] Zhang, Y., Niyato, D., Wang, P. and Han, Z. (2020) Data Services Sales Design with Mixed Bundling Strategy: A Multidimensional Adverse Selection Approach. IEEE Internet of Things Journal, 7, 8826-8836. [Google Scholar] [CrossRef] 
[14] Chapparadalli, N.L. (2019) Designing a Dashboard to Support the Decision Process of Dynamic Pricing. Master’s Thesis, University of Twente, Enschede.
[15] Wu, C., Buyya, R. and Ramamohanarao, K. (2019) Cloud Pricing Models: Taxonomy, Survey, and Interdisciplinary Challenges. ACM Computing Surveys (CSUR), 52, 1-36. [Google Scholar] [CrossRef]