Commodity demand forecasting using modulated rank reduction for humanitarian logistics planning

作者：

Highlights：

• Demand prediction for highly dynamic and short events.

• Application of robust principal component for data preparation for online training.

• We develop a method for optimizing training lag for online model training schemes.

• Real-world application for humanitarian supply chain demand prediction.

摘要

•Demand prediction for highly dynamic and short events.•Application of robust principal component for data preparation for online training.•We develop a method for optimizing training lag for online model training schemes.•Real-world application for humanitarian supply chain demand prediction.

论文关键词：Time-series prediction,Data efficient machine learning,Rank reduction,Walk-forward validation

论文评审过程：Received 29 November 2021, Revised 31 May 2022, Accepted 1 June 2022, Available online 9 June 2022, Version of Record 15 June 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.117753