Table 7.

Cross-validation results of the deep learning algorithms on phishtank dataset

Algorithm	Accuracy	AUC	F1-Score
DNN	91.13%	0.9125	0.9139
CNN	93.33%	0.9332	0.9349
RNN	97.22%	0.9720	0.9730
LSTM	98.23%	0.9822	0.9827
BiLSTM	98.27%	0.9826	0.9831
DNN+LSTM	98.98%	0.9901	0.9910
DNN+BiLSTM	99.21%	0.9934	0.9941

Bold values indicate the best overall result for the corresponding algorithm