Bio-Inspired Optimization Algorithm in Machine Learning and Practical Applications

Software fault prediction is critical to ensuring system reliability and user satisfaction while minimizing financial losses. This research paper presents a novel and comprehensive technique for predicting software faults by integrating a Genetic Algorithm (GA) with an ensemble KNN and XG Boost algorithms classifier. The proposed approach is evaluated on 24 datasets, including 12 open-source Java projects and 12 NASA Metrics Data Program datasets, with the NASA Metrics Data Program datasets serving as macroscale datasets and open-source Java projects datasets serving as microscale datasets. By optimizing the selection and combination of fault-related metrics, the integrated GA on the ensemble classifier demonstrates enhanced accuracy in fault prediction compared to existing methods. The experimental results highlight the effectiveness of the proposed approach, establishing its superiority in predictive accuracy. This study significantly contributes to software engineering, offering an efficient technique for identifying and mitigating potential software faults, thereby elevating software reliability and quality assurance processes.