A Comparative Study of Machine Learning Algorithms As an Audit Tool in Financial Failure Prediction

Authors : Şafak AĞDENİZ, Birol YILDIZ

Pages : 11-32

View : 18 | Download : 14

Publication Date : 2019-07-01

Article Type : Research Paper

Abstract :The main aim of this study is to show usage of machine learning as an audit tool. Within this main aim, the object of this study is to compare the classification performances of machine learning algorithms in financial failure and to determine the best algorithm. Financial failure has been one of the major research topic in accounting and finance. Financial failure is an important task for internal auditors too. As an assurance activity internal auditors should give an assurance about the company continuity Early studies used traditional statistical techniques. With the development of computer science and technology, artificial intelligence and machine learning have been used in order to increase the accuracy. The output that has been used in this study is classification accuracy. Our data set consist of 216 companies’ financial data between the period 1983-2012. As a result of the study it was seen that rule based classification algorithms’ are more successful than the others. The decision table algorithm from this rule based classification algorithms has reached the highest classification performance with a ratio of 91.8%.
Keywords : Financial Failure Prediction, Machine Learning, Classification, Data Mining, Audit Tool