Skip to main content
. Author manuscript; available in PMC: 2020 Jun 1.
Published in final edited form as: Comput Biol Med. 2019 Apr 25;109:85–90. doi: 10.1016/j.compbiomed.2019.04.018

Figure 2.

Figure 2.

Row 1, from left to right: AUCs of training from scratch for reduced VGGNet, using learning rate of 0.0001,0.0005 and 0.00002. Row 2, from left to right: AUCs of transfer learning using pre-trained GoogleNet, using learning rate 0.001,0.005 and 0.0002 for the last fully connected layer. Red line: test AUC; blue line: training AUC.