基于随机森林与费希尔判别法的区块链企业财务预警模型研究  

Financial Risk Forewarning Model of Block Chain Enterprise Based on Random Forest Algorithm and Fisher Discriminant Method

作　　者：余欣然 郭婷 Yu Xinran;Guo Ting(School of Management,Xiamen University,Xiamen 361000,Fujian,China;School of Economics and Management,Inner Mongolia University of Technology,Huhehaote 010000,Inner Mongolia,China)

机构地区：[1]厦门大学管理学院,福建厦门361000 [2]内蒙古工业大学经济管理学院,内蒙古呼和浩特010000

出　　处：《梧州学院学报》2022年第3期40-46,共7页Journal of Wuzhou University

摘　　要：近年来区块链技术的应用与发展成为热门,掀起了产业变革的狂潮。财务预警为企业防范及监控财务风险提供了理论依据,对企业的经营发展具有重要意义。以A股区块链企业为研究对象的前提下,采用随机森林算法提取特征指标并进一步结合费希尔判别法得出模型函数。结果显示:模型预测效果良好,基于随机森林的费希尔判别模型比单一的费希尔模型判别结果更加准确。该模型可为财务人员、企业管理者等提供财务风险判别依据。The application and development of block chain technology has been a hot issue,raising a surging tide of industry reform.Financial warning is a theoretical basis for enterprises prevention and financial risk monitoring,which is of great significance for the operation and development of enterprises.Focusing on A-share block chain enterprise,this paper extracts characteristic index through random forest algorithm and achieves model function by combining it with Fisher discriminant method.The result shows that the model reaches a good forecast effect and Fisher discriminant model based on random forest algorithm reaches a better effect than a single Fisher discriminant model.That model can be used by accounfants and enterprise managers to achieve more accurate determining criterion of financial risk.

关 键 词：区块链企业 财务预警 随机森林 费希尔判别法 

分 类 号：F275[经济管理—企业管理]