Skip to main content
. 2019 Aug 15;19(16):3556. doi: 10.3390/s19163556

Figure 5.

Figure 5

Figure 5

Comparative models accuracy and loss diagram. In (a) the VGG-16 model final accuracy is 97.83% for training and 98.86% for the validation. In (b) the final loss is 0.0572 for training and 0.042 on the validation set. In (c) the ReseNet-50 model final accuracy is 96.23% for training and 95.61% for the validation. In (d) the final loss is 0.103 for training and 0.102 on the validation set. In (e) the Inception model final accuracy is 96.77% for training and 95.42% for the validation. In (f) the final loss is 0.109 for training and 0.144 on the validation set.