随机森林模型在软件企业价值评估中的应用研究
Research on the Application of Stochastic Forest Model in the Value Evaluation of Software Enterprises
吴晓畅 1郭红霞1
作者信息
- 1. 广西科技大学 经济与管理学院,广西 柳州 545006
- 折叠
摘要
软件企业因其技术更新快、产品附加值高、易存未知性等特点,运用传统的评估方法容易造成其评估值失真.针对软件企业价值评估的难点与重点,文章提出了基于随机森林模型的软件企业价值评估方法,采用平均下降精度、平均下降基尼系数并结合相关系数聚类法构建模型指标体系.使用A股软件企业样本数据进行实证分析.研究得出:随机森林模型的拟合优度为 92.7%,拟合效果较好;在重要性分析上,影响软件企业价值最重要的三个因素为研发投入金额、营业总收入增长率和息前税后利润.
Abstract
Due to the characteristics of fast technology update,high added value of products,and easy to exist unknowns,software enterprises can easily cause distortion of their evaluation values by using traditional evaluation methods.In view of the difficulties and key points of the value evaluation of software enterprises,this paper proposes a software enterprise value evaluation method based on stochastic forest model,which adopts the average reduction accuracy,the average reduction of Gini coefficient and the correlation coefficient clustering method to build a model index system.It uses A-share software enterprise sample data for empirical analysis.The study shows that the optimal fit of stochastic forest model is 92.7%,and the fitting effect is good.In terms of importance analysis,the three most important factors affecting the value of software enterprises are the amount of research and development investment,the growth rate of total business revenue and net operating profit less adjusted taxes.
关键词
企业价值评估/软件企业/随机森林模型Key words
enterprise value assessment/software enterprises/stochastic forest model引用本文复制引用
基金项目
广西科技大学博士基金(03200184)
广西工业高质量发展中心项目(03220059)
出版年
2024
NETL
NSTL
维普
万方数据