software-bug-pred

Software bug prediction is the process of identifying software components that are more likely to have bugs in it. Bug prediction can be done by analyzing factors like bug reports and code metrics. This aims to develop a machine learning model to predict the number of bugs in advance. The goal of this project is identifying the bugs so that we will get a quality and bug free software.This project is about finding bugs in software bugs dataset. The dataset is taken from a software bug dataset website. It was cleaned for running the models using Python. The algorithms used in this project are machine learning algorithms. The algorithms used in this project are Random Forest, k-means and k nearest neighbour algorithm.The Dataset is a combination of five software systems datasets: Eclipse JDT Core, Mylyn, Lucene, Eclipse PDE UI and Equinox Framework. The combined dataset is a collection software metrics like number of lines, attributes, methods and bugs. There are 18 columns in the dataset. Column names: cbo, dit, fanIn, fanOut, Icom, noc, numberOfAttributes, numberOfAttributesInherited, numberOfLinesOfCode, numberOfMethods, numberOfMethodsInherited, numberOfPrivateAttributes, bugs etc. Number of instances: 5371 Number of attributes: 18