Detecting fraudulent financial statements: a data mining approach

One of the most important goals of an independent audit is to discover and identify fraudulent financial reports. By presenting a model that includes financial and non-financial criteria for discovering distorted financial statements, this research succeeded in identifying distorted financial reports in companies listed on the Tehran Stock Exchange using data mining techniques. The meaning of six variables from two financial and non-financial dimensions and using a sample of companies accepted in Tehran Stock Exchange consisting of ۱۳۰۳ years - companies (including ۲۱ fraudulent companies and ۱۶۸ non-fraudulent companies) during the years ۲۰۱۱ to ۲۰۲۱ have been investigated. and it has been analyzed using data mining techniques including decision tree, neural networks and Bayesian methods. The research results show that the Bayesian method more accurately discovers distorted financial statements.