Comparative Analysis of Techniques to Predict Fault Proneness

Software Quality and reliability are essential parts of software development process. Fault Proneness is a measure of data that can help the programmers to predict fault prone areas in the projects during testing or development process. This knowledge can prove very beneficial in improving software quality. Software Quality Estimation models can be broadly classified as classification and prediction. Classification techniques are used to predict probability of occurrence of fault but cannot be used to predict the number of faults. Whereas, count models such as the Poisson regression model, and the zero-inflated Poisson regression model can be used to obtain both a qualitative classification, and a quantitative prediction for software quality. In this paper we are reviewing models such as count and classification models to bring in light the most often used techniques by the researchers and academicians.