This study delves into the prediction of supply chain upstreamness using an ensemble machine learning approach. Leveraging insights from the analysis of trade credit and profitability in production networks, we develop a novel methodology to forecast the vertical position of firms within supply chains. These production networks are constructed based on supply chain relationships and accounting data from the FactSet and Compustat databases, with enterprise upstreamness correspondingly defined. By employing random forests, gradient boosting trees, and ensemble classifiers, and incorporating key variables such as various firm characteristics, our ensemble machine learning model aims to accurately predict the upstreamness of firms in complex production networks, demonstrating high accuracy and robustness. The findings shed light on the importance of upstreamness prediction methods for enterprises and offer valuable implications for supply chain management.
Predicting Supply Chain Upstreamness Using An Ensemble Machine Learning Method
23.08.2024
1292879 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Ensemble Machine Learning Classification Models for Predicting Pavement Condition
Transportation Research Record | 2024
|Predicting the Porosity of SCM-Blended Concrete Composites Using Ensemble Machine Learning Models
Springer Verlag | 2024
|Predicting Sector Complexity Using Machine Learning
AIAA | 2022
|Predicting Sector Complexity Using Machine Learning
TIBKAT | 2022
|