Research Article

Balanced Adversarial Tight Matching for Cross-Project Defect Prediction

Table 9

Ablation study on the metric of AUC.

Target projectNo-HFNo-MCNo-MDDBATM

Ant0.7810.7360.7750.801
Camel0.6530.6280.6390.678
Forrest0.3920.2830.4370.48
Ivy0.8560.7650.8170.869
Log4j0.5160.4470.4960.526
Lucene0.6380.6160.6390.672
Poi0.6760.6540.6630.705
Synapse0.7280.7060.7230.775
Velocity0.6480.6210.6350.736
Xalan0.6990.6870.7050.743
Xerces0.6180.5680.6010.689
Average0.6520.6110.6480.698
Improvement6.56%14.2%7.71%ā€”
p-Value0.0030.0030.003ā€”