Empirical study of Software Quality estimation
Inderpreet Kaur, Arvinder Kaur · 2012
Software Quality is an important nonfunctional requirement which is not satisfied by many software products. Prediction models using object oriented metrics can be used to identify the faulty classes. In this paper, we will empirically study the relationship between object oriented metrics and fault proneness of an open source project Emma. Twelve machine Learning classifiers have been used. Univariate and Multivariate analysis of Emma shows that Random Forest provides optimum values for accuracy, precision, sensitivity and specificity.