An Effort-Aware Just-in-Time Software Defect Prediction Model based on LightGBM

作者：Liqiong Chen, Ying Wang · 年份：2023 · DOI：10.1109/iciibms60103.2023.10347835 · 被引用次数：2 · 研究领域：Software Engineering Research、Software Reliability and Analysis Research、Software System Performance and Reliability

Effort-aware Just-in-Time Software Defect Prediction is a fine-grained technique and takes the cost of detection into account to detect more defective changes with limited testing resources. Since many existing models lack better performance, an effort-aware just-in-time software defect prediction model is proposed called EALGB based on LightGBM in this paper. Intensive simulation experiments have been conducted on six open-source datasets to demonstrate the performance of the proposed method. The prediction model is evaluated by two widely recognized metrics, ACC and Popt, in three scenarios called cross-validation, cross-project cross-validation and time-aware cross-validation. Empirical results demonstrate that the proposed algorithm EALGB has good prediction performance.