Table 5.

Performance of different classifiers in predicting protein ubiquitination using AAC, AAindex, and one-hot encoding for the Single-Type SBF Model.

Model (epochs = 30)	Loss (training)	Accuracy (validation)
DenseNet121	0.520	0.676
DenseNet201	0.581	0.683
ResNet34	0.162	0.688
Swin-base transformer	0.683	0.572
Swin tiny transformer	0.683	0.572
Vision-based transformer	0.670	0.590
Vision large transformer	0.694	0.562