基于时序矩阵分解的缺失销售数据估计

doi:10.12005/orms.2021.0356

运筹与管理 ›› 2021, Vol. 30 ›› Issue (11): 99-105.DOI: 10.12005/orms.2021.0356

基于时序矩阵分解的缺失销售数据估计

陈斯敏¹, 杨磊¹, 陈文娜², 黄晓宇¹

1.华南理工大学电子商务系,广东广州 510006;
2.华南理工大学经济与金融学院,广东广州 510006

收稿日期:2020-05-14 出版日期:2021-11-25
通讯作者: 黄晓宇(1977-),男,广东人,副教授,博士,主要研究方向为机器学习、统计与优化理论及应用。
作者简介:陈斯敏(1996-),女,广东人,硕士研究生,主要研究方向为数据驱动的供应链运作优化;杨磊(1979-),男,山东人,教授,博士,主要研究方向为供应链管理;陈文娜(1983-),女,广东汕头人,硕士研究生,主要研究方向为金融工程,运作优化。
基金资助:
华南理工大学中央高校基本科研业务费项目(XYMS202107);国家社科基金后期资助项目(20FGLB034);广州市人文社会科学重点研究基地成果(PZL2021KF0027);广东省自然科学基金项目(2019A1515010792、2020A1515010830);广东省哲学社科基金项目(GD17XYJ25)

Estimating the Missing Sales Data with the Time Series Matrix Factorization Model

CHEN Si-min¹, YANG Lei¹, CHEN Wen-na², HUANG Xiao-yu¹

1. Department of Electronic Business, South China University of Technology, Guangzhou 510006, China;
2. School of Economics and Finance, South China University of Technology, Guangzhou 510006, China

Received:2020-05-14 Online:2021-11-25

摘要/Abstract

摘要： 企业的历史销售记录是供应链优化研究的基础数据来源,然而,在日常的研究中,几乎所有可以通过公开途径获得的销售记录都是高度不完整的,这为研究者开展工作带来了极大的不便。为解决此问题,本文提出,以销售数据集中已有的数据为基础,使用面向时序数据的矩阵分解模型MAFTIS对其缺失的部分进行估算,从而把残缺的数据集补全完整。进一步地,为提高MAFTIS的计算效率,本文还为该模型设计了一种基于交替最小二乘法的求解策略MAFTIS^ALS。在评估实验中,MAFTIS^ALS被用于三个真实销售数据集的缺失记录估计,结果显示,与其它估计模型相比,MAFTIS^ALS能获得更准确的估计结果,并且具有更高的收敛速度。

关键词: 销售数据, 缺失值估计, 矩阵分解

Abstract: The historical sales data are one of the most important data sources for supply chain optimization (SCO) research. However, for most SCO researchers, the publicly accessible sales data is usually highly incomplete, which makes their work difficult. In order to solve this problem, we propose to use MAFTIS, atime series matrix factorization model, to estimate the missing sales data with respective to the collected ones.In addition, to speed up the computation procedure of MAFTIS, we also present an alternating least square (ALS) based solution algorithm for the model.For evaluations, we apply the proposed method to estimate the missing entries of three real sales data sets, and all results show that, compared with competitive algorithms, our proposed model achieve the best performance in terms of prediction accuracy and convergence speed.

Key words: sales data, missing value estimation, matrix factorization

中图分类号:

TP181

陈斯敏, 杨磊, 陈文娜, 黄晓宇. 基于时序矩阵分解的缺失销售数据估计[J]. 运筹与管理, 2021, 30(11): 99-105.

CHEN Si-min, YANG Lei, CHEN Wen-na, HUANG Xiao-yu. Estimating the Missing Sales Data with the Time Series Matrix Factorization Model[J]. Operations Research and Management Science, 2021, 30(11): 99-105.

参考文献

[1] Donselaar K H V, Peters J, Jong A D. Analysis and forecasting of demand during promotions for perishable items[J]. International Journal of Production Economics, 2016, 172(2): 65-75.
[2] Ma S, Fildes R, Huang T. Demand forecasting with high dimensional data: the case of SKU retail sales forecasting with intra-and inter-category promotional information[J]. European Journal of Operational Research, 2016, 249(1): 245-257.
[3] Loureiro A L D, Miguéis V L, da Silva Lucas F M. Exploring the use of deep neural networks for sales forecasting in fashion retail[J]. Decision Support Systems, 2018, 114(OCT.): 81-93.
[4] 徐琪,刘峥.基于SVM的短生命周期产品供应链双渠道需求预测模型[J].系统管理学报,2014,23(02):255-262,270.
[5] Haengju L, Yongsoon E. Discovering heterogeneous consumer groups from sales transaction data[J]. European Journal of Operational Research, 2020, 280: 338-350.
[6] Boone T, Ganshan R, Jain A, Sanders N R. Forecasting sales in the supply chain: consumer analytics in the big data era[J]. International Journal of Forecasting, 2019, 35: 170-180.
[7] Pivette M, Mueller J E, Crépey, Pascal, et al. Surveillance of gastrointestinal disease in france using drug sales data[J]. Epidemics, 2014, 8: 1-8.
[8] Arinaminpathy N, Batra D, Khaparde S, et al. The number of privately treated tuberculosis cases in india: an estimation from drug sales data[J]. The Lancet Infectious Diseases, 2016, 16: 1255-1260.
[9] Drewnowski A, Michels S, Leroy D. The impact of crunchy wednesdays, on happy meal fruit orders: analysis of sales data in france, 2009~2013[J]. Journal of Nutrition Education and Behavior, 2017, 49(3): 236-240.
[10] 侯小超,张磊,杨晴.基于蒙特卡罗方法的中国中长期煤炭需求预测[J].运筹与管理.2020,29(1):99-105.
[11] Sanders Nada R. Big data driven supply chain management: a framework for implementing analytics and turning information into intelligence[M]. Pearson Education, 2014.
[12] Ali M M, Babai M Z, Boylan J E, Syntetos A A. Supply chain forecasting when information is not shared[J]. European Journal of Operational Research, 2017, 260(3): 984-994.
[13] 黄晓宇,潘嵘,李磊,等.面向时序数据的矩阵分解[J].软件学报,2015,26(9):2262-2277.
[14] Yu W, Chavez R, Jacobs M A, Mengying F. Data-driven supply chain capabilities and performance: a resource-based view[J]. Transportation Research Part E: Logistics and Transportation Review, 2018(114): 371-385.
[15] Agrahri Himanshu, Ahmed F, Virendra Kumar Verma, Jayant K. Purohit. Benefits of implement big data driven supply chain management: an ISM based model[J]. International Journal of Engineering Science, 2017, 7(5): 11426-11431.
[16] 邱萍萍,黄晓宇,曾青松. 考虑边信息的多层贝叶斯需求预测模型[J].计算机集成制造系统,2020,26(第1):191-201.
[17] Croston J D. Forecasting and stock control for intermittent demands[J]. Journal of the Operational Research Society, 1972, 23(3): 289-303.
[18] Boylan J E, Syntetos A A. The accuracy of a modified croston procedure[J]. International journal of production economics, 2007, 107(2): 511-517.
[19] Syntetos A. Forecasting of intermittent demand[D]. Brunel University, 2001.
[20] Erik Levén, Segerstedt A. Inventory control with a modified croston procedure and erlang distribution[J]. International Journal of Production Economics, 2004, 90(3): 361-367.
[21] Teunter R H, Syntetos A A, Babai M Z. Intermittent demand: linking forecasting to inventory obsolescence[J]. European Journal of Operational Research, 2011, 214(3): 606-615.
[22] Nikolopoulos K I, Babai M Z, Bozos K. Forecasting supply chain sporadic demand with nearest neighbor approaches[J]. International Journal of Production Economics, 2016, 177: 139-148.
[23] Khashei M, Bijari M, Hejazi S R. Combining seasonal ARIMA models with computational intelligence techniques for time series forecasting[J]. Soft Computing, 2012, 16(6): 1091-1105.
[24] Van Ryzin G, Vulcano G. A market discovery algorithm to estimate a general class of nonparametric choice models[J]. Management Science, 2015, 61(2): 281-300.
[25] Chong E K P, Stanislaw H. Z·ak. An introduction to optimization, third edition[M]. Hoboken,JOHN WILKY & SONS, 2011.
[26] Golub Gene H, Charles F, Van Loan. Matrix computations[M]. Baltimore,Johns Hopkins University Press, 1996.

基于时序矩阵分解的缺失销售数据估计

Estimating the Missing Sales Data with the Time Series Matrix Factorization Model

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 3

编辑推荐

Metrics

[1]	高文欣, 刘升, 肖子雅. 闪电分叉过程算法优化的K-means聚类[J]. 运筹与管理, 2021, 30(12): 35-41.
[2]	孙卫红, 吕文新. 区域集群下板材订单配置模型及算法研究[J]. 运筹与管理, 2020, 29(12): 38-42.
[3]	胡崇海, 李一军, 姜维, 王铁军. 动态连续蚁群系统及其在天基预警中的应用[J]. 运筹与管理, 2011, 20(2): 89-96.