The Usefulness of Relief Feature Selection Method in Auditor's Opinion Type Prediction

Accomplished researches in auditor's opinion type prediction emphasize the design of more sophisticated classifiers, and less feature (variables) selection and their appropriate methods. The Purpose of this research is investigating the usefulness of relief method in auditor's opinion type prediction of companies listed on Tehran Stock Exchange. To obtain comparable experimental results, Support Vector Machine (SVM) and logistic regression were used in this study. In overall, the experimental results of investigating 1214 firms-years in 2008 to 2015, confirmed the usefulness and positive effects of relief Method in predicting auditors opinion type. In other words, the application of this feature selection method, increases the mean of accuracy, and reduces the occurrence of type I and type II errors. Furthermore, the results indicated that SVM outperforms the logistic regression.