Table 10 PofB20 Values of TR, CNN, DBN, LSTM, DP-HNN, SDP-BB, ACGDP, and CNN-MLP.

From: Semantic and traditional feature fusion for software defect prediction using hybrid deep learning model

Projects

TR

CNN

DBN

LSTM

DP-HNN

SDP-BB

ACGDP

CNN-MLP

Xalan-2.4

0.456

0.369

0.612

0.587

0.603

0.597

0.605

0.619

Xalan-2.6

0.512

0.435

0.552

0.611

0.580

0.616

0.582

0.626

Poi-1.5

0.436

0.563

0.574

0.550

0.569

0.577

0.584

0.566

Poi-3.0

0.532

0.563

0.473

0.487

0.592

0.482

0.533

0.492

Ant-1.7

0.241

0.221

0.296

0.361

0.305

0.351

0.366

0.406

log4j-1.1

0.376

0.386

0.305

0.372

0.322

0.374

0.320

0.415

jEdit-4.0

0.401

0.376

0.343

0.300

0.310

0.364

0.361

0.398

jEdit-4.1

0.388

0.371

0.334

0.339

0.377

0.371

0.377

0.389

Lucene-2.0

0.462

0.484

0.598

0.472

0.582

0.541

0.609

0.623

Lucene-2.2

0.453

0.514

0.615

0.624

0.605

0.636

0.611

0.612

Synapse-1.1

0.295

0.299

0.287

0.455

0.411

0.452

0.426

0.493

Synapse-1.2

0.285

0.274

0.292

0.480

0.436

0.467

0.488

0.521

Average

0.403

0.405

0.440

0.470

0.474

0.486

0.489

0.513