Detecting Financial Statement Fraud by Using The Integration of Random Forest and Rough Set

• Main Authors: Chia-Yi Chu, 朱家逸
• Other Authors: Der-Jang Chi
• Format: Others
• Language: zh-TW
• Published: 2010
• Online Access: http://ndltd.ncl.edu.tw/handle/52726053656446168496

碩士 === 中國文化大學 === 會計學系 === 98 === In these few years, the rapid change of global economic environment has increased the occurrence possibility of fraudulent financial statement. Therefore, to build up an appropriate fraudulent financial statement diagnosis model has become a very important task in industry. The objective of the proposed study is to investigate the performance of enterprise distress diagnosis by integrating the random forest with rough sets with discriminant analysis technique. In addition to the financial capital indicator, the corporate governance indicator is also included in the model to measure the assets of companies. The results from the present study indicate that the proposed combined approach predict much accurate and coverage much faster than that the rough sets. Moreover, we find out that the diagnostic correctness of enterprise distress is significantly influenced by both financial indicators and corporate governance indicators.