.png){kind=icon}
.png){kind=icon}
Return
A novel software defect prediction model using two-phase grey wolf optimisation for feature selection
.png){kind=icon}
PRE
AI
DOI：10.1007/s10586-024-04599-w.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
The process of accurately predicting software defects is highly crucial during the early period of software development before testing activities begin. A variety of computational methods have been constructed to achieve this based on static code metrics. However, one of the major issues in predictive modelling is the presence of redundant and irrelevant features in available datasets, which can lead to inaccuracies in the prediction model. Swarm optimization methods have shown excellent performance in Feature Selection (FS) issue mitigation and reduced the execution time of the prediction model. This study proposes a novel model for predicting software defects. This model utilizes a variant of Grey Wolf Optimiser as a wrapper-based feature selection method, paired with Synthetic Minority Oversampling Technique to balance the dataset, with the objective of maximizing the prediction efficiency of the learning model. The performance of the proposed model is assessed on 27 open-source datasets. The result findings show that the feature selection method improves prediction performance. Furthermore, the two-phase Grey Wolf Optimization-based feature selection with Random Forest classifier demonstrates superior efficacy on datasets compared to another benchmark model in handling the problem of FS. The results are also validated using statistical techniques.
Keywords:
Software defect prediction
Feature selection
Grey wolf optimisation
Synthetic minority oversampling technique (SMOTE)
Machine learning
Journal
C
IF:
4.1
Papers:
5.0K
Citations:
7.5K