基于大数据的金融风险预警系统的研究  

A Research on Early-warning System for Financial Risks Based on Big Data

作　　者：卢民荣[1,2] 甘健胜 Minrong;Gan Jiansheng(School of Accountancy,Fujian Jiangxia University,Fuzhou,Fujian 350108,China;Finance and Accounting Research Center of Fujian Philosophy Social Science Research Base,Fuzhou,Fujian 350108,China)

机构地区：[1]福建江夏学院会计学院,福建福州350108 [2]福建省社科研究基地财务与会计研究中心,福建福州350108

出　　处：《湖北工程学院学报》2020年第2期86-96,共11页Journal of Hubei Engineering University

基　　金：福建省社科基地重大项目(FJ2018JDZ014、FJ2019JDZ053);福建省中青年教师科研项目重点课题(JZ180190);福建省本科高校教育教学改革研究重大项目(FBJG20190295)。

摘　　要：历史大数据显示,政治、军事、金融、疫情事件是金融风险触发的重大因素,网民、投资者对事件的关注度越大,事件影响力随之扩大,从而把金融事件传播转向金融风险传染。通过用户量级、用户结构、用户行为、用户指数构建用户参与模型,并以此模型构建金融风险预警系统。系统通过采集2000年至2018年大量相关支撑数据,在建立风险指标、用户参与等各类数据库基础上,运用机器学习算法设计预警、预测系统,最终用系统采集的数据库进行效果分析和相关性检验,得出量化结论。Historical big data demonstrate that political,financial and military events and also disease outbreak prove to be the major factors that may cause financial risks.The more net-citizens and investors pay attention to the events,the more influence the events may exert and become financial risk inflection.Early-warning system for financial risks is constructed on the basis of users’involvement mode which is established via users’levels,structure,behaviors and indicator.The system collects a large amount of related supporting data from 2000 to 2018 to make an analysis of the effect and test the correlation and employs machine learning algorithm to design early-warning and predicting system based on risks indicator,users’involvement and various database.

关 键 词：大数据 金融风险 预警系统 风险传染 重大事件 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]