Efficient Fault Prediction Using Exploratory and Causal Techniques

Software is basically a series or cluster of operational directions or instructions. The operations or functions performed by the system are regulated by a set of orderly arranged instructions. An error in the code of the developed software is called as software fault. This field attracted many researchers to work in this domain not only due to its advantages but also due to availability of open source dataset and existence of a lot of research publications in the domain. The study performs exploratory and causal relation technique between metrics and bugs. Exploratory Factor analysis is used to identify the important variables of bugs. The identified variables are used to develop a robust model. This study is the extension of our previous experiment in which some variables were analyzed to determine the important predictors. And these distinct predictors were identified and a robust regression model was developed. In this study, we used the same model but development and identification mechanism of variables is different. The results prove the capability of the technique used. The comparison of results presented in the study. On the basis of results obtained researchers are provided with future guidelines in this research work.