Defect Prediction Model for AOP-based Software Development using Hybrid Fuzzy C-Means with Genetic Algorithm and K-Nearest Neighbors Classifier

Abstract

The number of defects has often been considered a vital indicator of quality of software. It is well known that we cannot go back and add quality. Software Quality and reliability are considered to be one of the most important concerns of software product. In this paper, we give a brief overview of an Aspect-Oriented Programming (AOP) and a model is proposed to predict defects. The model is empirically validated on the PROMISE Software Engineering Repository dataset with three different types of methods. One is Fuzzy CMeans Clustering (FCM) approach and another is K-Nearest Neighbors (KNN) classifier technique, have been performed in real data. Third is a hybrid approach (i.e. combination of fuzzy c-means and genetic algorithms) have been performed. The performance of data is evaluated in terms of Reliability, Accuracy, Mean Absolute Error (MAE), and Root Mean Square Error (RMSE). General Terms AOP, AOSD