REPD: Source Code Defect Prediction As Anomaly Detection (CROSBI ID 681854)
Prilog sa skupa u zborniku | izvorni znanstveni rad | međunarodna recenzija
Podaci o odgovornosti
Afrić, Petar ; Šikić, Lucija ; Kurdija, Adrian Satja ; Delač, Goran ; Šilić, Marin
engleski
REPD: Source Code Defect Prediction As Anomaly Detection
In this paper, we present a novel approach to defect prediction within project source code. Since defect prediction datasets are typically imbalanced, and there are few defective examples, we treat defect prediction as anomaly detection. We present our Reconstruction Error Probability Distribution (REPD) model and compare it on five different datasets to five standardly used models: Gaussian Naive Bayes, Logistic regression, k- nearest-neighbors, decision tree, and SVM. For the main performance results we use F1-scores. Using statistical means, we show that our model produces significantly better results, improving F1-score up to 10.11%.
Defect prediction ; Program analysis ; Binary classification
nije evidentirano
nije evidentirano
nije evidentirano
nije evidentirano
nije evidentirano
nije evidentirano
Podaci o prilogu
227-234.
2019.
objavljeno
10.1109/QRS-C.2019.00052
Podaci o matičnoj publikaciji
Podaci o skupu
2019 IEEE 19th International Conference on Software Quality, Reliability and Security Companion (QRS-C)
predavanje
22.07.2019-26.07.2019
Sofija, Bugarska