RESEARCH ON FRAUD FINANCIAL STATEMENTS APPROACH BY RANDOM FOREST ALGORITHM

  • Đặng Ngọc Hùng
  • Phạm Thị Hồng Diệp
  • Cao Thị Nhiên
Keywords: Fraud; financial report; machine learning; fraud triangle; random forest

Abstract

Research using random forest algorithms based on fraudulent triangle
theory, to examine the influence of factors on fraudulent financial statements of
enterprises in Vietnam, using a dataset Table data with 2235 observations for the
period 2014 - 2020. The study considers the difference in profit before and after
the audit over 5%. Forecast results according to random forest algorithm reach
91%. At the same time, the study has determined that the most influential factor
is the financial stability of enterprises. This study is valuable when it comes to
assessing the financial reporting fraud situation of enterprises with the ability to
predict with high accuracy.

điểm /   đánh giá
Published
2022-06-02
Section
KINH TẾ-XÃ HỘI