Deep Learning-Based Efficient Model Development for Phishing Detection Using Random Forest and BLSTM Classifiers

Table 2

Training parameters of the BLSTM network.

BLSTM architecture	Hidden layers
	h1	h2	h3	h4	h5
Training instances	245	245	245	245	245
Validating instances	100	100	100	100	100
Learning rate	0.001	0.0001	0.0001	0.01	0.0001
Activation function	ReLU	ReLU	ReLU	ReLU	ReLU
Number of epochs	200	600	800	900	1000
Training time (s)	130	150	160	165	200
Accuracy (%)	81.97	89.02	95.87	95.88	94.30
AUC (%)	81.27	88.52	93.47	95.68	94.53