Table 8 Comparison of proposed work with recent existing techniques on the HAM10000 dataset.

From: Skin lesion classification of dermoscopic images using machine learning and convolutional neural network

Existing work versus proposed work

Accuracy (%)

Proposed CNN

95.18

EW-FCM+wide-shufflenet29

84.80

Shifted MobileNetV230

81.90

Shifted GoogLeNet30

80.50

Shifted 2-Nets30

83.20

MobileNetV2-LSTM8

85.34

24 layered CNN22

86.50

InceptionV331

91.56

ResNetXt10131

93.20

InceptionResNetV231

93.20

Xception31

91.47

NASNetLarge31

91.11

9 layered CNN32

80.00

Resnet50, Restnet101+KcPCA+SVM RBF21

89.80

Vgg16+googLeNet ensemble33

81.50

Modified MobileNet34

83.93

Back to article page