International Journal of Advanced Research in Computer Science & Technology (IJARCST)
Software reliability is becoming an important part of any software system. It is an important factor in software quality since it calibrates software failures. Clustering is a process of partitioning a group of data objects into a small number of clusters on the basis of similarity. K-nearest neighbor is a clustering method that aims to find the defects in the new software based on the previously developed software. In this paper, the authors will examine the defects of the software with the help of k-means algorithm. Finally, they propose a defect prediction system with high performance which manages to decrease the cost of software system simultaneously.