Skip to main content
. 2021 May 7;34(3):605–617. doi: 10.1007/s10278-021-00455-0

Table 4.

Training time of each fold using RF and GBRT. GBRT needed longer time to train (1068 s on average) compared to RF (20.8 s on average) (KF method was used for splitting the dataset and augmented images were used for training)

Training time per fold (s)
RF GBRT
Fold 0 22 1039
Fold 1 21 1084
Fold 2 21 1111
Fold 3 20 1054
Fold 4 20 1053
Mean ± std 20.8 ± 0.7 1068.2 ± 25.9