Skip to main content
. 2021 Sep 5;186:115805. doi: 10.1016/j.eswa.2021.115805

Table 13.

Summary of the best achieved Top-1 combinations in all experiments.

Experiment Model Parameters optimizer Batch size Dropout ratio TL learn ratio WSM
1 VGG16 SGD 32 0.58 80% 99.02%
2 VGG19 SGD 32 0.56 80% 99.31%
3 ResNet50 SGD 32 0.6 100% 97.64%
4 ResNet101 SGD 64 0.6 100% 96.63%
5 DenseNet121 RMSProp 32 0.41 75% 87.96%
6 DenseNet169 AdaMax 32 0.36 50% 88.54%
7 MobileNet SGD 64 0.6 100% 97.44%
8 MobileNetV2 AdaDelta 32 0.28 0% 63.19%
9 Xception AdaMax 64 0.6 100% 97.29%