Cross-Project Defect Prediction Based on Two-Phase Feature Importance Amplification
Table 3
F1-measure, AUC, and MCC of models with different classifications, including Naive Bayes (NB), Logistic Regression (LR), Decision Tree (DT), Support Vector Classification (SVC), and Random Forest (RF).
F1-measure
AUC
MCC
LR
DT
NB
SVC
RF
LR
DT
NB
SVC
RF
LR
DT
NB
SVC
RF
EQ-JDT
0.594
0.729
0.560
0.571
0.790
0.758
0.852
0.713
0.727
0.873
0.479
0.655
0.467
0.465
0.735
EQ-LC
0.435
0.619
0.404
0.224
0.761
0.675
0.889
0.654
0.575
0.903
0.387
0.601
0.361
0.136
0.738
EQ-ML
0.375
0.672
0.333
0.376
0.742
0.647
0.866
0.611
0.637
0.875
0.271
0.627
0.253
0.288
0.702
EQ-PDE
0.379
0.660
0.361
0.386
0.776
0.664
0.851
0.621
0.667
0.864
0.259
0.607
0.283
0.269
0.740
JDT-EQ
0.595
0.829
0.492
0.459
0.833
0.680
0.857
0.644
0.632
0.862
0.379
0.727
0.358
0.347
0.719
JDT-LC
0.362
0.573
0.373
0.337
0.752
0.694
0.837
0.672
0.687
0.875
0.293
0.539
0.303
0.266
0.726
JDT-ML
0.376
0.637
0.361
0.418
0.745
0.644
0.854
0.632
0.679
0.883
0.276
0.589
0.265
0.320
0.707
JDT-PDE
0.373
0.651
0.356
0.381
0.761
0.668
0.841
0.619
0.679
0.866
0.253
0.595
0.279
0.265
0.721
LC-EQ
0.672
0.806
0.578
0.729
0.831
0.727
0.839
0.678
0.773
0.862
0.453
0.665
0.389
0.542
0.710
LC-JDT
0.575
0.721
0.585
0.525
0.816
0.755
0.860
0.743
0.730
0.911
0.451
0.648
0.473
0.384
0.768
LC-ML
0.393
0.655
0.350
0.406
0.765
0.647
0.843
0.620
0.659
0.855
0.307
0.604
0.272
0.314
0.732
LC-PDE
0.394
0.660
0.367
0.378
0.787
0.675
0.843
0.628
0.665
0.866
0.279
0.605
0.278
0.258
0.755
ML-EQ
0.696
0.826
0.560
0.632
0.855
0.730
0.857
0.674
0.687
0.886
0.455
0.703
0.397
0.369
0.755
ML-JDT
0.567
0.725
0.580
0.539
0.787
0.748
0.855
0.728
0.742
0.883
0.440
0.650
0.483
0.404
0.730
ML-LC
0.343
0.597
0.422
0.338
0.824
0.686
0.889
0.681
0.656
0.925
0.273
0.583
0.363
0.261
0.806
ML-PDE
0.401
0.653
0.381
0.395
0.763
0.684
0.837
0.635
0.680
0.865
0.288
0.597
0.295
0.281
0.724
PDE-EQ
0.640
0.791
0.497
0.525
0.873
0.704
0.826
0.645
0.638
0.897
0.410
0.643
0.357
0.299
0.785
PDE-JDT
0.585
0.701
0.602
0.521
0.806
0.746
0.834
0.752
0.733
0.881
0.470
0.619
0.495
0.382
0.754
PDE-LC
0.382
0.584
0.438
0.362
0.784
0.702
0.857
0.690
0.639
0.873
0.315
0.557
0.380
0.306
0.763
PDE-ML
0.385
0.629
0.312
0.406
0.754
0.646
0.831
0.600
0.666
0.853
0.291
0.573
0.232
0.308
0.718
Average
0.476
0.686
0.446
0.445
0.790
0.694
0.851
0.662
0.678
0.878
0.351
0.619
0.349
0.323
0.739
The data in bold shows the classification with the best perfomance in each set of experiments.
