A Study on the Detection of Fraudulent Financial Statements
In order to improve the detection method of fraud financial statements, this paper divided 60 fraud companies into three types according to their industry and characteristics.Different identification indicators were selected for different types of companies.Whats more, the indicators based on Benfords law were included in the fraud detection indicator system.The recursive feature elimination algorithm and support vector machine were used to perform feature selection and model training, constructing fraud detection models for three types of companies.The result shows that the three fraud detection models contain different identification indicators, and their recognition accuracy for test samples reach 100%, 93%, and 73%, respectively.
financial statement fraud Benfords law sample classification support vector machine
Daibin Xiao Haiqi Feng
Central University of Finance and Economics Beijing,China
国际会议
郑州
英文
425-430
2018-09-21（万方平台首次上网日期，不代表论文的发表时间）