基于需求不确定性的数据驱动库存管理研究综述

doi:10.3969/j.issn.1007-7375.230259

摘要/Abstract

摘要： 随着高质量数据的日益丰富、机器学习技术的持续进步以及计算能力的显著提升，数据驱动库存管理正迎来前所未有的发展机遇。然而，目前学术界对于这一新兴领域的研究进展尚缺乏全面系统的综述。本研究运用文献计量方法，深入分析了183篇学术论文，并通过科学知识图谱的可视化方式，全面展示了该领域的研究现状。从大数据和运营管理的双重视角出发，总结归纳了数据驱动库存管理在需求信息、基本模型和基本方法3个方面的研究结果。重点从需求不确定性和特征数据的角度介绍了4种库存管理模型：单变量数据驱动报童模型、单变量数据驱动动态库存模型、多特征数据驱动报童模型和多特征数据驱动动态库存模型。在此基础上，梳理了6种主要的数据驱动决策方法，包括贝叶斯方法、鲁棒优化方法、样本均值近似方法、分位数回归方法、操作统计方法和机器学习方法。最后，本研究从数据驱动库存管理方法与工具层面，以及面临的难点与应用热点层面，提出了未来研究的方向与建议，旨在为相关领域的研究者和实践者提供有益的参考和启示，推动数据驱动库存管理领域不断发展。

关键词: 数据驱动, 库存管理, 报童, 动态库存, 研究综述

Abstract: In recent years, with the increasing abundance of high-quality data, continuous development of machine learning techniques and significant improvements of computational capabilities, data-driven inventory management is experiencing unprecedented development opportunities. However, comprehensive and systematic reviews of research advances in this emerging field are currently lacking. In this study, an in-depth analysis of 183 academic papers is conducted using bibliometrics, and the state of the art in this field is visualized through scientific knowledge graphs. Then, the research results of data-driven inventory management from the perspectives of big data and operation management are summarized and synthesized in three aspects: demand information, basic models and basic methods. Essentially, this paper introduces four inventory management models from the perspectives of demand uncertainty and feature data: univariate data-driven newsvendor model, univariate data-driven dynamic inventory model, multi-feature data-driven newsvendor model and multi-feature data-driven dynamic inventory model. On this basis, six main data-driven decision-making methods are summarized: Bayesian analysis, robust optimization, sample average approximation, quantile regression, operation statistics and machine learning. Finally, future research directions and suggestions are discussed from the perspectives of methodologies, tools, challenges, and application hotspots in data-driven inventory management, aiming to provide valuable references and insights for researchers and practitioners in the relevant fields, and to foster the continuous development of data-driven inventory management.

Key words: data-driven, inventory management, newsvendor, dynamic inventory, literature review

中图分类号:

邵思淇, 钟远光, 陈植, 李延希. 基于需求不确定性的数据驱动库存管理研究综述[J]. 工业工程, 2024, 27(3): 1-11.

SHAO Siqi, ZHONG Yuanguang, CHEN Zhi, LI Yanxi. A Review of Data-driven Inventory Management Based on Demand Uncertainty[J]. Industrial Engineering Journal, 2024, 27(3): 1-11.

参考文献

[1] DING X, PUTERMAN M L, BISI A. The censored newsvendor and the optimal acquisition of information[J]. Operations Research, 2002, 50(3): 517-527
[2] QIN Y, WANG R, VAKHARIA A J, et al. The newsvendor problem: review and directions for future research[J]. European Journal of Operational Research, 2011, 213(2): 361-374
[3] HUBER J, MÜLLER S, FLEISCHMANN M, et al. A data-driven newsvendor problem: from data to decision[J]. European Journal of Operational Research, 2019, 278(3): 904-915
[4] CORLU C G, BILLER B, TAYUR S. Driving inventory system simulations with limited demand data: insights from the newsvendor problem[J]. Journal of Simulation, 2019, 13(2): 152-162
[5] PRAK D, TEUNTER R. A general method for addressing forecasting uncertainty in inventory models[J]. International Journal of Forecasting, 2019, 35(1): 224-238
[6] SCARF H E. Some remarks on Bayes solutions to the inventory problem[J]. Naval Research Logistics Quarterly, 1960, 7(4): 591-596
[7] AZOURY K S. Bayes solution to dynamic inventory models under unknown demand distribution[J]. Management Science, 1985, 31(9): 1150-1160
[8] SCARF H E. A min-max solution of an inventory problem[M]. Santa Monica: Rand Corporation, 1957.
[9] PERAKIS G, ROELS G. Regret in the newsvendor model with partial information[J]. Operations Research, 2008, 56(1): 188-203
[10] LEVI R, PERAKIS G, UICHANCO J. The data-driven newsvendor problem: new bounds and insights[J]. Operations Research, 2015, 63(6): 1294-1306
[11] TAKEUCHI I, LE Q V, SEARS T D, et al. Nonparametric quantile estimation[J]. Journal of Machine Learning Research, 2006, 7: 1231-1264
[12] NAHMIAS S. Demand estimation in lost sales inventory systems[J]. Naval Research Logistics, 1994, 41(6): 739-757
[13] AGRAWAL N, SMITH S A. Estimating negative binomial demand for retail inventory management with unobservable lost sales[J]. Naval Research Logistics, 1996, 43(6): 839-861
[14] LU X, SONG J, ZHU K. Analysis of perishable-inventory systems with censored demand data[J]. Operations Research, 2008, 56(4): 1034-1038
[15] HUH W T, RUSMEVICHIENTONG P. A nonparametric asymptotic analysis of inventory planning with censored demand[J]. Mathematics of Operations Research, 2009, 34(1): 103-123
[16] HUH W T, LEVI R, RUSMEVICHIENTONG P, et al. Adaptive data-driven inventory control with censored demand based on Kaplan-Meier estimator[J]. Operations Research, 2011, 59(4): 929-941
[17] SACHS A L, MINNER S. The data-driven newsvendor with censored demand observations[J]. International Journal of Production Economics, 2014, 149: 28-36
[18] LARIVIERE M A, PORTEUS E L. Stalking information: Bayesian inventory management with unobserved lost sales[J]. Management Science, 1999, 45(3): 346-363
[19] LEVI R, ROUNDY R O, SHMOYS D B. Provably near-optimal sampling-based policies for stochastic inventory control models[J]. Mathematics of Operation Research, 2007, 32(4): 821-839.
[20] ZHENG M, WU K, SHU Y. Newsvendor problems with demand forecast updating and supply constraints[J]. Computers & Operations Research, 2016, 67: 193-206
[21] GODFREY G A, POWELL W B. An adaptive, distribution-free algorithm for the newsvendor problem with censored demands, with applications to inventory and distribution[J]. Management Science, 2001, 47(8): 1101-1112
[22] CARRIZOSA E, OLIVARES-NADAL A V, RAMÍREZ-COBO P. Robust newsvendor problem with autoregressive demand[J]. Computers & Operations Research, 2016, 68: 123-133
[23] TAUBE F, MINNER S. Data-driven assignment of delivery patterns with handling effort considerations in retail[J]. Computers & Operations Research, 2018, 100: 379-393
[24] ÇIMEN M, KIRKBRIDE C. Approximate dynamic programming algorithms for multidimensional flexible production-inventory problems[J]. International Journal of Production Research, 2017, 55(7): 2034-2050
[25] AKCAY A, CORLU C G. Simulation of inventory systems with unknown input models: a data-driven approach[J]. International Journal of Production Research, 2017, 55(19): 5826-5840
[26] GIANNOCCARO I, PONTRANDOLFO P. Inventory management in supply chains: a reinforcement learning approach[J]. International Journal of Production Economics, 2002, 78(2): 153-161
[27] HEKIMOĞLU M, VAN DER LAAN E, DEKKER R. Markov-modulated analysis of a spare parts system with random lead times and disruption risks[J]. European Journal of Operational Research, 2018, 269(3): 909-922
[28] KARA A, DOGAN I. Reinforcement learning approaches for specifying ordering policies of perishable inventory systems[J]. Expert Systems with Applications, 2018, 91: 150-158
[29] LIU J, PANG Z, QI L. Dynamic pricing and inventory management with demand learning: a Bayesian approach[J]. Computers & Operations Research, 2020, 124: 105078
[30] PEREZ H D, HUBBS C D, LI C, et al. Algorithmic approaches to inventory management optimization[J]. Processes, 2021, 9(1): 102
[31] BAN G Y, RUDIN C. The big data newsvendor: practical insights from machine learning[J]. Operations Research, 2019, 67(1): 90-108
[32] BEUTEL A L, MINNER S. Safety stock planning under causal demand forecasting[J]. International Journal of Production Economics, 2012, 140(2): 637-645
[33] OROOJLOOYJADID A, SNYDER L V, TAKÁČ M. Applying deep learning to the newsvendor problem[J]. IISE Transactions, 2020, 52(4): 444-463
[34] ZHANG Y, GAO J. Assessing the performance of deep learning algorithms for newsvendor problem[C]//Neural Information Processing: 24th, International Conference, ICONIP 2017. Cham: Springer, 2017: 912-921.
[35] BERTSIMAS D, KALLUS N. From predictive to prescriptive analytics[J]. Management Science, 2020, 66(3): 1025-1044
[36] BERTSIMAS D, MCCORD C. From predictions to prescriptions in multistage optimization problems[EB/OL]. (2019-04-26)[2023-12-26]. https://arxiv.org/abs/1904.11637.
[37] CLAUSEN J B B, LI H. Big data driven order-up-to level model: application of machine learning[J]. Computers & Operations Research, 2022, 139: 105641
[38] CHEUNG W C, SIMCHI-LEVI D. Sampling-based approximation schemes for capacitated stochastic inventory control models[J]. Mathematics of Operations Research, 2019, 44(2): 668-692
[39] CAO Y, SHEN Z. Quantile forecasting and data-driven inventory management under nonstationary demand[J]. Operations Research Letters, 2019, 47(6): 465-472
[40] QIU R, SUN Y, FAN Z, et al. Robust multi-product inventory optimization under support vector clustering-based data-driven demand uncertainty set[J]. Soft Computing, 2020, 24(9): 6259-6275
[41] TAYLOR J W. A quantile regression approach to estimating the distribution of multiperiod returns[J]. The Journal of Derivatives, 1999, 7(1): 64-78
[42] ZHANG L, YANG J, GAO R. Optimal robust policy for feature-based newsvendor[J]. Management Science, 2024, 70(4): 2315-2329.
[43] KARLIN S. Dynamic inventory policy with varying stochastic demands[J]. Management Science, 1960, 6(3): 231-258
[44] IGLEHART D L. The dynamic inventory problem with unknown demand distribution[J]. Management Science, 1964, 10(3): 429-440
[45] LOVEJOY W S. Myopic policies for some inventory models with uncertain demand distributions[J]. Management Science, 1990, 36(6): 724-738
[46] BERTSIMAS D, THIELE A. A robust optimization approach to inventory theory[J]. Operations Research, 2006, 54(1): 150-168
[47] GEORGHIOU A, TSOUKALAS A, WIESEMANN W. Robust dual dynamic programming[J]. Operations Research, 2019, 67(3): 813-830
[48] 孙月, 邱若臻. 交叉销售下基于支持向量聚类的数据驱动多产品库存鲁棒优化模型[J]. 中国管理科学, 2022, 30(2): 156-168
SUN Yue, QIU Ruozhen. Support vector clustering-based data driven robust optimization model for multi-product inventory problem with cross-selling[J]. Chinese Journal of Management Science, 2022, 30(2): 156-168
[49] 邱若臻, 苑红涛, 黄小原. 基于似然估计的零售商库存鲁棒均值–风险模型[J]. 中国管理科学, 2016, 24 (8) : 123-131.
QIU Ruozhen, YUAN Hongtao, HUANG Xiaoyuan. Robust mean-risk model for retailer inventory problem based on likelihood estimation[J]. Chinese Journal of Management Science, 2016, 24 (8) : 123-131.
[50] CHEN Z, XIE W. Regret in the newsvendor model with demand and yield randomness[J]. Production and Operations Management, 2021, 30(11): 4176-4197
[51] LU M, SHEN Z. A review of robust operations management under model uncertainty[J]. Production and Operations Management, 2021, 30(6): 1927-1943
[52] KLEYWEGT A J, SHAPIRO A, HOMEM-DE-MELLO T. The sample average approximation method for stochastic discrete optimization[J]. SIAM Journal on Optimization, 2002, 12(2): 479-502
[53] HALMAN N. Provably near-optimal approximation schemes for implicit stochastic and sample-based dynamic programs[J]. INFORMS Journal on Computing, 2020, 32(4): 1157-1181
[54] CANNON A J. Quantile regression neural networks: implementation in R and application to precipitation downscaling[J]. Computers & Geosciences, 2011, 37(9): 1277-1284
[55] NGUYEN T T, NGUYEN N D, NAHAVANDI S. Deep reinforcement learning for multiagent systems: a review of challenges, solutions, and applications[J]. IEEE Transactions on Cybernetics, 2020, 50(9): 3826-3839
[56] WANG J, LI X, ZHU X. Intelligent dynamic control of stochastic economic lot scheduling by agent-based reinforcement learning[J]. International Journal of Production Research, 2012, 50(16): 4381-4395
[57] ZHANG Y, VOVK V, ZHANG W. Probability-free solutions to the non-stationary newsvendor problem[J]. Annals of Operations Research, 2014, 223(1): 433-449
[58] WANG R, GAN X, QIU T, et al. An extended weak aggregating algorithm for a two dimensional data-driven multi-stage newsvendor problem[J]. Expert Systems with Applications, 2024, 238: 121992
[59] BOOKBINDER J H, LORDAHL A E. Estimation of inventory re-order levels using the bootstrap statistical procedure[J]. IIE Transactions, 1989, 21(4): 302-312
[60] BURNETAS A N, SMITH C E. Adaptive ordering and pricing for perishable products[J]. Operations Research, 2000, 48(3): 436-443
[61] POWELL W, RUSZCZYŃSKI A, TOPALOGLU H. Learning algorithms for separable approximations of discrete stochastic optimization problems[J]. Mathematics of Operations Research, 2004, 29(4): 814-836