Financial Fraud Identification of Listed Companies Based on Support Vector Machine and Lasso-Logistic Regression
220 companies punished for financial fraud from 2010 to 2023 are selected as the research objects.First,a financial fraud identification model is constructed based on the support vector machine algorithm.On this basis,Lasso regression is used to analyze the core factors affecting financial fraud,and Logisitc regression is used to further analyze the different degrees of influence of financial fraud.It is found that asset-liability ratio,age and investment cash flow have a positive impact on the probability of financial fraud,while the net profit rate of total assets has a negative impact.Therefore,listed companies should focus on and control these indicators when improving their internal control system.At the same time,regulators should also pay attention to these indicators and be good at using fraud models to improve the identification effect of financial fraud.The state should actively promote the formation of information sharing mechanism a-mong various regulatory agencies,and constantly update the input variables of fraud models in order to obtain better forecasting results.
万方数据
维普
