Different Feature Selection Methods and Auditor's Opinion Type Prediction

This research is aimed to investigate and compare the effects of different feature selection methods on auditor's opinion type prediction. To do so, this research compares the performance of feature selection methods (including correlation-based, t-test, stepwise discriminate analysis, relief and factor analysis). Two classifiers used in this study are support vector machine and neural networks. The sample includes 1214 firms-years listed in the Tehran Stock Exchange in the period from 2008 to 2015. The results confirm the effectiveness of feature selection methods and significant differences among their performance. In other wordsc, using the feature selection methods increases the mean of accuracy and reduces type I and type II errors. Furthermore, the results indicate that support vector machine outperforms the neural networks.