摘要
本文基于制造业企业面临的转型环境特征以及风险管理框架中的风险源评估要素,提出研发型制造业的风险测度指标体系,并利用主成分分析法进一步筛选关键指标,利用随机森林算法构建基于机器学习的研发型制造业财务风险预警模型。研究发现,该财务风险预警模型预测准确率较高,总资产净利润率、财务费用率和研发投入强度等是预测研发型企业财务风险的关键指标。该研究结果对处于数字化转型阶段的研发型制造业企业实施有效的风险管理具有实际的应用价值。
This article proposes a risk measurement indicator system for R&D driven manufacturing based on the characteristics of the transforming environment faced by manufacturing enterprises and the risk source assessment elements in the risk management framework.Principal component analysis is used to further screen key indicators,and a machine learning based financial risk warning model for R&D manufacturing is constructed using random forest algorithm.This paper finds that the financial risk warning model has high prediction accuracy,and the key indicators for predicting financial risks in R&D driven enterprises include total asset net profit margin,financial expense ratio,and R&D expense ratio.The research results have practical application value for implementing effective risk management for R&D manufacturing enterprises in the digital transformation stage.
作者
柯剑
王琦琦
Ke Jian;Wang Qiqi
出处
《管理会计研究》
2023年第5期22-32,共11页
MANAGEMENT ACCOUNTING STUDIES
关键词
研发型制造业
财务风险
机器学习
R&D Driven Manufacturing Industry
Financial Risks
Machine Learning