Table 8 Comparison of the hybrid model with various classifier

From: A comprehensive analysis of deep learning and transfer learning techniques for skin cancer classification

Pretrained model

ML Classifier

TP

FN

FP

TN

Sensitivity

Precision

Accuracy

F-score

ResNet-18 + VGG-19

SVM

330

42

30

258

0.8871

0.91667

0.89091

0.90164

DT

262

77

98

223

0.77286

0.72778

0.73485

0.74964

KNN

310

55

50

245

0.84932

0.86111

0.84091

0.85517

Naïve bayes

270

79

90

221

0.77364

0.7500

0.74394

0.76164

VGG-19 + MobileNet V2

SVM

339

30

21

270

0.91869

0.94167

0.92272

0.93001

DT

282

92

78

208

0.75401

0.78333

0.74242

0.76839

KNN

325

75

35

225

0.8125

0.90278

0.83333

0.85526

Naïve bayes

233

88

127

212

0.72586

0.64722

0.67424

0.68429

MobileNet V2 + ResNet-18

SVM

340

20

27

273

0.94444

0.92643

0.92878

0.93534

DT

287

70

77

230

0.8017

0.78611

0.77727

0.79383

KNN

319

65

41

235

0.83073

0.88611

0.83939

0.85753

Naïve bayes

225

73

135

224

0.75503

0.6250

0.75667

0.68389

Back to article page