As software projects become more complex, there is an increased focus on their analysis and testing. Detecting software faults is a problem of major importance for improving the quality of the software development related processes and the efficiency of the software testing process. In order to detect faults in existing software systems, we introduce in this paper a novel approach, based on fuzzy self-organizing feature maps. A fuzzy map will be trained, using unsupervised learning, to provide a two-dimensional representation of the faulty and non-faulty entities from a software system and it will be able to identify if a software module is or not a defective one. Five open-source case studies are used for the experimental evaluation of our approach. The obtained results are better than most of the results already reported in the literature for the considered datasets and emphasize that a fuzzy self-organizing map is more efficient than a crisp one for the case studies used for evaluation.
Software defect detection, Machine learning, Self-organizing map, Fuzzy clustering.
Istvan-Gergely CZIBULA, Gabriela CZIBULA, Zsuzsanna MARIAN, Vlad-Sebastian IONESCU, "A Novel Approach Using Fuzzy Self-Organizing Maps for Detecting Software Faults", Studies in Informatics and Control, ISSN 1220-1766, vol. 25(2), pp. 207-216, 2016. https://doi.org/10.24846/v25i2y201608
All articles published by Studies in Informatics and Control (SIC) are published in accordance with the terms and conditions of the Creative Commons Attribution License (CC BY-NC-SA 4.0).