Data Mining Methods for Fault Prediction & Clustering & Classification

In Software fault detection, clustering and classification are major issues in software engineering the software engineering deals with various techniques of prediction such as correction cost, fault, test, effort, reusability, security and quality prediction etc. A most common and widespread area of research is software fault prediction. It is termed as the method of developing models which are utilized by the software specialists for identifying faulty constructs in the prior levels of Software Development Life Cycle. Classical and modern methods have been used in this regard. But some intelligent computing methods have been found very useful for the clustering of faults such as ANN. GA. KBS, ES and Data mining methods. In the thesis we have given an extensive research literature survey for the various intelligent computing methods as mentioned above. Also basic concepts on ANN Clustering, C&RT are given. Our main contribution to the work is to deploy data mining methods for clustering and classification of software faults. WE have also determined the importance different types of metrics of faults.