Using Chinese listed company to detect and analyze the fraudulent financial reporting by data mining technology

• Main Authors: HSU,HSUN-HAO, 許巡豪
• Other Authors: Ma,Chia-Ying
• Format: Others
• Language: zh-TW
• Published: 2018
• Online Access: http://ndltd.ncl.edu.tw/handle/h2522y

碩士 === 東吳大學 === 會計學系 === 106 === With the reform of the Chinese mainland in 1978, socialism with capitalism brought considerable economic growth to China, but it also faced a huge information asymmetry in the Chinese market. Some proprietor usually used this information asymmetry to engage in fraud, which exposes investors to great investment risks and exposes accounting firms to huge audit costs. This study uses the Bayesian belief network, neural network, support vector machine and decision tree in Logis regression and data exploration to analyze the fraudulent behaviors listed in China. The research concludes 29 frauds from 2009 to 2016. The company and its 58 non-fraud companies. The results show that the neural network has higher predictive power regardless of the full sample or fraudulent behavior.