Software Bug Prediction Using Machine Learning Approach

Abstract

Abstract: Software defect prophecy work focuses on the number of faults remaining in a software system. A prophecy of the number of remaining scars in an audited artefact can be used for decision timber. An accurate prophecy of the number of scars in a software product during system testing contributes not only to the operation of the system testing process but also to the estimation of the product’s demanded conservation. amiss software modules beget software failures, increase development and conservation costs, and drop customer satisfaction. It strives to meliorate software quality and testing effectiveness by constructing predictive models from law attributes to enable a timely identification of fault-prone modules. In this design, we will explore machine knowledge ways for software defect prophecy . This helps the formulators to descry software scars and correct them. Unsupervised ways may be used for defect prophecy in software modules, more so in those cases where defect labels are not available.