Welcome to the InfoSci Platform
Reference Hub6
Software Defect Prediction Using Hybrid Distribution Base Balance Instance Selection and Radial Basis Function Classifier

Software Defect Prediction Using Hybrid Distribution Base Balance Instance Selection and Radial Basis Function Classifier

Mrutyunjaya Panda
Copyright: © 2019 |Volume: 8 |Issue: 3 |Pages: 23
ISSN: 2160-9772|EISSN: 2160-9799|EISBN13: 9781522568049|DOI: 10.4018/IJSDA.2019070103
Cite Article Cite Article

MLA

Panda, Mrutyunjaya. "Software Defect Prediction Using Hybrid Distribution Base Balance Instance Selection and Radial Basis Function Classifier." IJSDA vol.8, no.3 2019: pp.53-75. http://doi.org/10.4018/IJSDA.2019070103

APA

Panda, M. (2019). Software Defect Prediction Using Hybrid Distribution Base Balance Instance Selection and Radial Basis Function Classifier. International Journal of System Dynamics Applications (IJSDA), 8(3), 53-75. http://doi.org/10.4018/IJSDA.2019070103

Chicago

Panda, Mrutyunjaya. "Software Defect Prediction Using Hybrid Distribution Base Balance Instance Selection and Radial Basis Function Classifier," International Journal of System Dynamics Applications (IJSDA) 8, no.3: 53-75. http://doi.org/10.4018/IJSDA.2019070103

Export Reference

Mendeley
Favorite Full-Issue Download

Abstract

Software is an important part of human life and with the rapid development of software engineering the demands for software to be reliable with low defects is increasingly pressing. The building of a software defect prediction model is proposed in this article by using various software metrics with publicly available historical software defect datasets collected from several projects. Such a prediction model can enable the software engineers to take proactive actions in enhancing software quality from the early stages of the software development cycle. This article introduces a hybrid classification method (DBBRBF) by combining distribution base balance (DBB) based instance selection and radial basis function (RBF) neural network classifier to obtain the best prediction compared to the existing research. The experimental results with post-hoc statistical significance tests shows the effectiveness of the proposed approach.

Request Access

You do not own this content. Please login to recommend this title to your institution's librarian or purchase it from the IGI Global bookstore.