Software Bug Prediction using KNN and Naïve Base algorithms

Abstract

The drift towards technological use has made the software an essential solution for many systems. The software's predictions and defect identification are one of the current research areas needed. A software malfunction results in an unintended or incorrect outcome that may be due to a bug, a fault, an error, or a mistake. In the early stages of software development this has to be detected. Otherwise, this could result in major risk factors such as wasting time, quality, resources, costs and time. Software defects may be identified early using data mining techniques. Two supervised learning techniques are applied to the data, such as the k-nearest neighbor and the Naïve based algorithm, to predict the classifier defects and accuracy.