Research on artificial intelligence-assisted software defect prediction and repair techniques
Yin Zheng
Liaoning University of Science and Technology
DOI: https://doi.org/10.59429/esta.v12i1.9680
Keywords: Artificial intelligence; Software defect prediction; Bayesian networks; Causal analysis; Software defect repair that
Abstract
In this paper, the application of artificial intelligence technology in software defect prediction and repair is studied with respect to the shortcomings of traditional software defect prediction models in terms of interpretability and robustness. Focusing on the modeling method of Bayesian network in the field of software defect prediction, a software defect prediction model based on Bayesian network is established through data discretization, Bayesian network structure learning algorithm, and probabilistic inference technology. Further, this paper integrates Bayesian network with common predictors such as K-nearest neighbor, decision tree, logistic regression and so on in a soft-voting manner to construct an integrated software defect prediction model. Experimental simulations are carried out on six publicly available software defect datasets, and the results show that the integrated model based on Bayesian network significantly outperforms the traditional integrated model in terms of evaluation indexes such as F1, Recall, and G-Mean.
References
[1] Interpretability-Driven Sample Selection Using Self Supervised Learning for Disease Classification and Segmentation.[J]. Mahapatra Dwarikanath;Poellinger Alexander;Shao Ling;Reyes Mauricio.IEEE transactions on medical imaging.2021.23-24.
[2] Towards Visual Explainable Active Learning for Zero-Shot Classification.[J]. Jia Shichao;Li Zeyu;Chen Nuo;Zhang Jiawan.IEEE transactions on visualization and computer graphics.2021.45-48.
[3] LAL: Meta-Active Learning-based Software Defect Prediction[J]. Yubin Qu;;Fang Li;;Xiang Chen.International Journal of Performability Engineering.2023.56-57.
[4] ALTRA: Cross-Project Software Defect Prediction via Active Learning and Tradaboost[J]. Yuan Zhidan;Chen Xiang;Cui Zhanqi;Mu Yanzhou.IEEE Access.2024.67-68.