摘要
随着我国资本市场的不断壮大,财务造假和市场波动等问题不断出现。为了准确衡量企业价值,设计了基于长期价值的统计指标企业评估模型。对比分析发现,采用改进随机森林算法的模型效果更好,平衡数据集训练能显著提升预测性能;预测分析发现,现金流是最重要的特征。长期价值平衡模型的各项指数显著提升,增益率达到0.87。综上所述,该模型具备较高的应用能力。
As China's capital market continues to grow,problems such as financial fraud and market volatility emerge continuously.In order to accurately measure the enterprise value,this paper designs a statistical indicator enterprise assessment model based on long-term value.Through comparative analysis,it is found that the model with improved random forest algorithm works better,and the balanced dataset training can significantly improve the prediction performance.After the prediction analysis,it is found that cash flow is the most important feature.The various indices of the long-term value balanced model improve significantly,and the gain rate reaches 0.87.The results show that the model has high application capability.
作者
周清明
彭涛
ZHOU Qingming;PENG Tao(School of Economics and Business,Hunan University of Technology,Zhuzhou 412007,China)
出处
《长春大学学报》
2023年第11期29-33,共5页
Journal of Changchun University
基金
湖北省教育厅项目(D20233011)。
关键词
长期价值
财务指标
随机森林算法
long-term value
financial indicators
fandom forest algorithm