A Comparison of Naïve Bayes and Random Forest for Software Defect Prediction

Abstract

The software defect can cause the unnecessary effects on the software such as cost and quality. The prediction of the software defect can be useful for the developing of good quality software. For the prediction, the PROMISE public dataset will be used and Random Forest (RF) algorithm and Naïve Bayes algorithm (NB) will be applied with the RAPIDMINER machine learning tool. This paper will compare the performance evaluation upon the different number of trees in RF and NB. As the results, the accuracy will be slightly increased if the number of trees will be more in RF. The maximum accuracy is up to 99.59 for RF and 97.12 for NB. The minimum accuracy is 87.13 RF and 45.87 for NB. Another comparison is based on AUC and all of the results show that RF algorithm is more accurate than Naïve Bayes algorithm for this defect prediction.