Open AccessBest Suited Machine Learning Techniques for Software Fault Prediction
Author(s) -
S Devika,
P L Lekshmy
Publication year - 2020
Publication title -
international journal of recent technology and engineering
Language(s) - English
Resource type - Journals
ISSN - 2277-3878
DOI - 10.35940/ijrte.f9456.038620
Subject(s) - software fault tolerance , software quality , computer science , software construction , software , fault (geology) , verification and validation , software metric , software system , machine learning , software development , reliability engineering , identification (biology) , software engineering , artificial intelligence , engineering , operating system , operations management , botany , seismology , biology , geology
In this world of emerging applications of software, it is always important to provide a quality assured product to customers. Software Fault Prediction popularly abbreviated as SFP is a major field which helps to provide quality assured products to customers. It helps to recognize modules that are bug- free and bug-prone in a software module. Machine learning techniques for both classification and determination are used for the purpose of software fault prediction. Software Fault Prediction is carried out prior to testing process without executing the source code, instead vital characteristics of software is taken into consideration. This early identification of faults can help software engineers to reduce the risk of system failure. A company does not always prefer to invest more expense on testing and in those situations, software fault prediction can have an upper hand in testing. The software fault prediction model will first train the learning techniques to generate base learners and then apply these base learners to unseen projects. It is always preferred to determine the count of faults rather than classifying each software module as fault-free and fault-prone. All software fault prediction techniques depend on base learners used and also nature of fault dataset. In this paper, the major learning techniques to determine software fault, characteristics of software fault dataset, etc. are discussed.