A Systematic Review on Inventory Forecasting Using Deep Learning

Abstract

Recent years have seen an upsurge in using deep learning in inventory forecasting. It enables firms to create exact estimates regarding the demand for their goods. Deep learning for inventory forecasting might result in data-driven decision-making and enhanced supply chain operations. A comprehensive assessment is necessary for advancing the industry and ensuring the success of using these technologies. A systematic review provides a forum for comparing and assessing various procedures and approaches, allowing for discovering the most promising and practical solutions. It could also highlight additional research areas, such as developing more sophisticated deep-learning algorithms and incorporating other data sources. There was one article published in 2018, but none in 2019. Six publications were released yearly in 2020 and 2021. As the industry expands, innovative research is expected. China and the US publish the most. Computer science and engineering dominate these articles. 11% of articles cover corporate management and accounting, highlighting the importance of understanding deep learning's impact on business. The exhaustive analysis using the Scopus search engine, search keywords, operators, and filters, and the Critical Appraisal Skills Programme (CASP) tool has certain limitations. Since the tool's foundation was the reviewers' judgements on the quality of the articles, the CASP tool is probably susceptible to bias. Agriculture and biology, the arts and humanities, ecology, medicine, psychology, and sociology are all understudied subjects that should be the focus of future research. As a field with the potential to have a significant influence, it is crucial to investigate how deep learning revolutionises business management and accounting.

Keywords: Inventory Forecasting, Deep Learning, Data-Driven Decision Making, Systematic Review

Read the article