Article: Design of manufacturing enterprise FEW system based on ML from the perspective of circular economy Journal: International Journal of Critical Infrastructures (IJCIS) 2025 Vol.21 No.9 pp.1 - 20 Abstract: This study designs a financial early warning system for manufacturing enterprises, focusing on machine learning and the circular economy. The random forest model is used as the base model, optimised by the artificial jellyfish algorithm to enhance prediction accuracy. Financial and non-financial indicators are selected through significance testing and feature screening methods. The results show that the optimised model achieves the highest accuracy of 88.42% and AUC of 0.918. Key warning indicators include inventory turnover rate, accounts receivable turnover rate, Herfindahl index, and liquidity ratio. The study highlights the importance of timely warnings for maintaining financial stability in manufacturing enterprises, helping them manage financial crises and supporting sustainable growth. The proposed system provides valuable support for policymakers and industry leaders in managing financial risks and advancing circular economy goals. Inderscience Publishers - linking academia, business and industry through research

Title: Design of manufacturing enterprise FEW system based on ML from the perspective of circular economy

Authors: Yang Huang; Xinyu Li; Dan Li

Addresses: Wuhan Maritime Communication Research Institute, Wuhan 430205, China ' Wuhan Maritime Communication Research Institute, Wuhan 430205, China ' Faculty of Management, Wuhan Donghu College, Wuhan 430212, China

Abstract: This study designs a financial early warning system for manufacturing enterprises, focusing on machine learning and the circular economy. The random forest model is used as the base model, optimised by the artificial jellyfish algorithm to enhance prediction accuracy. Financial and non-financial indicators are selected through significance testing and feature screening methods. The results show that the optimised model achieves the highest accuracy of 88.42% and AUC of 0.918. Key warning indicators include inventory turnover rate, accounts receivable turnover rate, Herfindahl index, and liquidity ratio. The study highlights the importance of timely warnings for maintaining financial stability in manufacturing enterprises, helping them manage financial crises and supporting sustainable growth. The proposed system provides valuable support for policymakers and industry leaders in managing financial risks and advancing circular economy goals.

Keywords: circular economy; CE; manufacturing enterprises; financial early warning; FEW; random forest; RF; artificial jellyfish algorithm.

DOI: 10.1504/IJCIS.2025.148646

International Journal of Critical Infrastructures, 2025 Vol.21 No.9, pp.1 - 20

Received: 15 Oct 2024
Accepted: 10 Feb 2025
Published online: 17 Sep 2025 *

Title: Design of manufacturing enterprise FEW system based on ML from the perspective of circular economy

Keep up-to-date