Skip to main content

Table 3 Training and validation accuracy with and without dropout layer

From: Accuracy of advanced deep learning with tensorflow and keras for classifying teeth developmental stages in digital panoramic imaging

 

Training accuracy

Validation accuracy

Without dropout

97.74

96.63

With dropout (p = 0.1)

96.17

96.17

With dropout (p = 0.5)

86.45

91.42

With dropout (p = 0.7)

78.60

90.66

With dropout (p = 0.9)

52.45

73.35

Back to article page

BMC Medical Imaging

ISSN: 1471-2342

Contact us