Pooling and Boosting for Demand Prediction in Retail: A Transfer Learning Approach

Forthcoming, Manufacturing & Service Operations Management

44 Pages Posted: 26 Jun 2023 Last revised: 23 Apr 2024

See all articles by Dazhou Lei

Dazhou Lei

Tsinghua University - Department of Industrial Engineering

Yongzhi Qi

JD.com Smart Supply Chain Y

Sheng Liu

Rotman School of Management

Dongyang Geng

JD.com Smart Supply Chain Y

Jianshen Zhang

JD.com Smart Supply Chain Y

Hao Hu

JD.com Smart Supply Chain Y

Zuo-Jun Max Shen

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

Date Written: September 10, 2022

Abstract

How should retailers leverage aggregate (category) sales information for individual product demand prediction? Motivated by inventory risk pooling, we develop a new prediction framework that integrates category-product sales information to exploit the benefit of pooling. We propose to combine data from different aggregation levels in a transfer learning framework. Our approach treats the top-level sales information as a regularization for fitting the bottom-level prediction model. We characterize the error performance of our model in linear cases and demonstrate the benefit of pooling. Moreover, our approach exploits a natural connection to regularized gradient boosting trees that enable a scalable implementation for large-scale applications. Based on an internal study with JD.com on more than 6,000 weekly observations between 2020 and 2021, we evaluate the out-of-sample forecasting performance of our approach against state-of-the-art benchmarks. The result shows that our approach delivers superior forecasting performance consistently with more than 9% improvement over the benchmark method of JD.com. We further validate its generalizability on a Walmart retail data set and through alternative pooling and prediction methods. Using aggregate sales information directly may not help with product demand prediction. Our result highlights the value of transfer learning to demand prediction in retail with both theoretical and empirical support. Based on a conservative estimate of JD.com, the improved forecasts can reduce the operating cost by 0.01-0.29 RMB per sold unit on the retail platform, which implies significant cost savings for the low-margin e-retail business.

Keywords: demand prediction, retail, machine learning, boosting trees

Suggested Citation

Lei, Dazhou and Qi, Yongzhi and Liu, Sheng and Geng, Dongyang and Zhang, Jianshen and Hu, Hao and Shen, Zuo-Jun Max, Pooling and Boosting for Demand Prediction in Retail: A Transfer Learning Approach (September 10, 2022). Forthcoming, Manufacturing & Service Operations Management, Available at SSRN: https://ssrn.com/abstract=4490516 or http://dx.doi.org/10.2139/ssrn.4490516

Dazhou Lei

Tsinghua University - Department of Industrial Engineering ( email )

Beijing
China

Yongzhi Qi

JD.com Smart Supply Chain Y ( email )

Sheng Liu (Contact Author)

Rotman School of Management ( email )

105 St. George st
Toronto, ON M5S 3E6
Canada

Dongyang Geng

JD.com Smart Supply Chain Y ( email )

Jianshen Zhang

JD.com Smart Supply Chain Y ( email )

Hao Hu

JD.com Smart Supply Chain Y ( email )

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
267
Abstract Views
704
Rank
215,347
PlumX Metrics