Skip to main content
. 2019 Jul 1;26(11):1286–1296. doi: 10.1093/jamia/ocz089

Figure 5.

Figure 5.

Boxplot of the classification performance of various models and depths. (A) The classification performance of three models using depth 16; (B) the performance of three models using depth 32; (C) the performance of three models using depth of 64. Models 3 and 4 (Convolutional AutoEncoder based Semi-supervised Network) achieve similar prediction performance and consistently outperforms model 2. Model 4 (Convolutional AutoEncoder based Semi-supervised Network) at depth 32 achieves the best average performance (0.824  ±  0.0329). The asterisk indicates an average value and a red circle indicates outliers. ***P < .005. Ave.: average; Med.: median; n.s.: not significant.