Cross-Project Defect Prediction Using Transfer Learning with Long Short-Term Memory Networks

Table 4

f-measure values for each model in CPDP.

Target project	LR	NNFilter	TCA	DBN	DPDBN	DPCNN	DPTCNN	TLSTM	DPTLSTM
Camel-1.6	0.328	0.325	0.324	0.31	0.336	0.341	0.331	0.309	0.322
Forrest-0.8	0.192	0.165	0.114	0.112	0.161	0.148	0.133	0.16	0.127
Log4j-1.2	0.644	0.649	0.669	0.66	0.657	0.68	0.684	0.689	0.652
Synapse-1.2	0.511	0.504	0.517	0.451	0.491	0.511	0.502	0.47	0.532
Xalan-2.7	0.609	0.611	0.66	0.661	0.642	0.653	0.655	0.675	0.64
Xerces-1.4.4	0.643	0.617	0.611	0.556	0.57	0.645	0.659	0.676	0.693

The values in bold represent the most optimal values in a row of data.