Skip to main content
. 2019 Dec 27;2:25. doi: 10.1186/s42492-019-0032-7

Table 7.

AUC scores of LMA without duplicate angles over 100 training and testing groups

Angles = 13 Angles = 49 Angles = 145
S = 1 0.8818 ± 0.0438 0.8833 ± 0.0387 0.8926 ± 0.0383
S = 2 0.8621 ± 0.0487 0.8524 ± 0.0512 0.8515 ± 0.0499
S = 3 0.9043 ± 0.0382 0.9054 ± 0.0388 0.9062 ± 0.0387
S = 4 0.8919 ± 0.0382 0.9084 ± 0.0376 0.8918 ± 0.0385
S = 5 0.9159 ± 0.0408 0.9176 ± 0.0337 0.9263 ± 0.0369
S = 6 0.9057 ± 0.0394 0.9077 ± 0.0377 0.9121 ± 0.0383
S = 7 0.9450 ± 0.0285 0.9401 ± 0.0319 0.9388 ± 0.0285
S = 8 0.8887 ± 0.0366 0.8962 ± 0.0335 0.8939 ± 0.0362
S = 9 08825 ± 0.0452 0.9226 ± 0.0364 0.9202 ± 0.0384
Average 0.8975 ± 0.0399 0.9037 ± 0.0377 0.9037 ± 0.0382

AUC Area under the curve of receiver operating characteristic curve, LMA Learning model by multiple angles