Risk Screening Of Listed Companies' Financial Statements Fraud

With the expansion of international capital market,international financial innovation and the rapid development of international financial business,China,as the second largest economy in the world,has gradually improved its market economic system and increased its total economic output;with the rise of information industry economy in the Internet environment,and with the financial market serving economic entities.With the continuous innovation and development of business model and financial derivatives,the business operation mode of real economy is becoming more and more complex and diversified.With the change of international environment and operating conditions under the economic globalization,the marketization of financial information has attracted more and more attention of relevant subjects.The fraudulent means of financial statements have reached a new height,showing a high level,high intelligence quotient,high innovation,high technology and high complexity,which brings great risks to the smooth operation of the national economy.By using modern statistical techniques,we can study various financial fraudulent practices,so as to prevent the occurrence of systematic funds at present.National macro-control and management of financial risk is of great significance.In order to prevent and forecast the fraud of financial statements in a forward-looking way,this paper collects the sample data of financial statements of 307 A-share listed companies from 2013 to 2017,chooses 10 core financial variable indicators,carries out in-depth and systematic statistical analysis and Research on the fraud risk screening of financial statements,and puts forward financial statements.Solutions to the problem of binary classification of report fraud detection.By establishing discriminant analysis model,logistic regression model,support vector machine model,decision tree model and neural network model,the sample data are forecasted by SPSS and IBM SPSS Modeler tools,and the forecasting results are compared and analyzed.The results show that the comprehensive forecasting accuracy of the neural network model is as high as 87.1%.It has a good forecasting performance for the fraud risk screening of financial statements.We innovatively construct the comprehensive early warning index model and calculate the critical value of the early warning index of financial fraud.We can supervise the fraud risk of financial statements beforehand,monitor in the event and take preventive measures afterwards,and effectively warn the hidden risk of financial statements fraud.