Comparing performances and effectiveness of machine learning classifiers in detecting financial accounting fraud for Turkish SMEs

Abstract

Turkish small-and medium-sized enterprises (SMEs) are exposed to fraud risks and creditor banks are facing big challenges to deal with financial accounting fraud. This study explores effectiveness of machine learning classifiers in detecting financial accounting fraud assessing financial statements of 341 Turkish SMEs from 2013 to 2017. The data are obtained from one of the leading creditor banks of Turkey. Highly imbalanced classes of 1384 nonfraudulent cases and 321 fraudulent cases (by 122 firms) are detected thus sampling techniques are used to mitigate class imbalance problem. Research methodology consists of two stages. First stage is data preprocessing wherein financial ratio calculation, feature selection methods for defining financial ratios with the greatest impact on fraudulent financial statements and two sampling methods of Synthetic Minority Oversampling Technique (SMOTE) as oversampling and undersampling are performed, respectively. Second stage is performance evaluation and comparison of classifiers wherein seven different classifiers (support vector machine, Naive Bayes, artificial neural network, K-nearest neighbor, random forest, logistic regression, and bagging) are executed and compared by using performance metrics. Classifiers are also compared without using any feature selection and/or sampling techniques. Results reveal that random forestwithout feature selection-oversampling model outperforms all other models.

Predicting Fraudulent Financial Statement Risk: The Testing Dechow F-Score Financial Sector Company Inindonesia

“…Besides, the limited audit process period becomes an obstacle in the audit process [20]. A study [21,22] shows that clients can use the auditor switching mechanism to reduce the possibility of observe fraud in the company's financial statements.…”

“…New auditors need a lot of time to study the company's characteristics in the audit process, so they are not able to know in detail the potential for fraud of the company. The research of [21] revealed that audit failure increased due to the change of auditors in the company. Study [18,20,37,38] shows that clients can use the auditor switching mechanism to reduce the prospect of detecting fraudulent acts in the company's financial reporting.…”

“…Arrogance is an attitude of primacy or greed from people who believe that internal control does not apply personally [39]. According to [21,23], the number of images of the Chief Executive Officer (C.E.O.) displayed in the company's annual finance report can represent the level of arrogance or primacy of the C.E.O.…”

Detection of Financial Statement Fraud Using Deep Learning for Sustainable Development of Capital Markets under Information Asymmetry

“…The traditional approaches to the detection of financial statement fraud are regression analysis, discriminant analysis, cluster analysis, and factor analysis. Hamal and Senvar [34] suggest that the detection of financial statement fraud requires sophisticated analytical tools and techniques, rather than the traditional methods adopted by decision-makers like auditors. However, there is no one-size-fits-all method for the detection of financial statement fraud according to the literature.…”

“…Moreover, data mining and machine learning techniques do not need the presumptions required by traditional statistics and can effectively handle non-linear problems [9]. For example, methods such as Artificial Neural Network (ANN), Support Vector Machine (SVM), Decision Tree (DT), and Bayesian Belief Network (BBN) are being used to detect financial statement fraud [7,8,10,28,32,34,38,40,41]. In the era of artificial intelligence, deep learning techniques are used by studies for the detection of financial statement fraud [35,42].…”