A Practical End-to-End Inventory Management Model with Deep Learning

27 Pages Posted: 23 Feb 2021 Last revised: 12 Jul 2021

See all articles by Meng Qi

Meng Qi

University of California, Berkeley - Department of Industrial Engineering and Operations Research

Yuanyuan Shi

Department of Electrical and Computer Engineering, University California, San Diego

Yongzhi Qi

JD.com Smart Supply Chain Y

Chenxin Ma

JD.com JD.com Silicon Valley Research Center

Rong Yuan

JD.com JD.com Silicon Valley Research Center

Di Wu

JD.com American Technologies Corporation

Zuo-Jun Max Shen

University of California, Berkeley - Department of Industrial Engineering & Operations Research (IEOR)

Date Written: November 25, 2020

Abstract

We investigate a data-driven multi-period inventory replenishment problem with uncertain demand and vendor lead time (VLT), with accessibility to a large quantity of historical data. Different from the traditional two-step predict-then-optimize (PTO) solution framework, we propose a one-step end-to-end (E2E) framework that uses deep-learning models to output the suggested replenishment amount directly from input features without any intermediate step. The E2E model is trained to capture the behavior of the optimal dynamic programming solution under historical observations, without any prior assumptions on the distributions of the demand and the VLT. By conducting a series of thorough numerical experiments using real data from one of the leading e-commerce companies, we demonstrate the advantages of the proposed E2E model over conventional PTO frameworks.
We also conduct a field experiment with JD.com and the results show that our new algorithm reduces holding cost, stockout cost, total inventory cost and turnover rate substantially compared to JD's current practice.
For the supply-chain management industry, our E2E model shortens the decision process and provides an automatic inventory management solution with the possibility to generalize and scale. The concept of E2E, which uses the input information directly for the ultimate goal, can also be useful in practice for other supply-chain management circumstances.

Keywords: end-to-end, inventory management, deep learning

Suggested Citation

Qi, Meng and Shi, Yuanyuan and Qi, Yongzhi and Ma, Chenxin and Yuan, Rong and Wu, Di and Shen, Zuo-Jun Max, A Practical End-to-End Inventory Management Model with Deep Learning (November 25, 2020). Available at SSRN: https://ssrn.com/abstract=3737780 or http://dx.doi.org/10.2139/ssrn.3737780

Meng Qi (Contact Author)

University of California, Berkeley - Department of Industrial Engineering and Operations Research ( email )

4141 Etcheverry Hall
Berkeley, CA 94720-1777
United States

Yuanyuan Shi

Department of Electrical and Computer Engineering, University California, San Diego ( email )

Yongzhi Qi

JD.com Smart Supply Chain Y ( email )

Chenxin Ma

JD.com JD.com Silicon Valley Research Center ( email )

Rong Yuan

JD.com JD.com Silicon Valley Research Center ( email )

Di Wu

JD.com American Technologies Corporation ( email )

United States

Zuo-Jun Max Shen

University of California, Berkeley - Department of Industrial Engineering & Operations Research (IEOR) ( email )

IEOR Department
4135 Etcheverry Hall
Berkeley, CA 94720
United States

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
583
Abstract Views
1,320
rank
57,526
PlumX Metrics