Predicting Auditor’s Opinions: A Neural Networks Approach

Abstract

Data mining methods can be used in order to facilitate auditors to issue their opinion. This paper initially applies two Data Mining classification techniques to develop models capable of identifying auditor’s opinion in Iran. The techniques used are Multilayer Perceptron neural network and Logistic regression. The period of this research is start of 2003 to end of 2009. The input vector is compose of financial data such as financial distress and non-financial data such as firm litigation. The four developed models are compared in terms of their performance. The results demonstrate the high explanatory power of the MLP model in identification of audit opinion. The model developed is accurate in classifying the total sample correctly with rate 87/75%. The model is also found to outperform traditional logistic regression. The result of this study can be useful to internal and external auditors, Investors, creditors, company decision-makers and other stakeholders.

Keywords

• auditor’s opinion
• multilayer perceptron neural networks
• data mining
• logistic regression