Predicting Pulsars from Imbalanced Dataset with Hybrid Resampling Approach

Table 14

Architecture of LSTM, DNN, and GRU used for experiments.

LSTM	DNN	GRU
LSTM (32)	Dense (64, activation = “relu”)	GRU (64, return_sequences = True))
Dropout (0.2)	Dropout (0.2)	SimpleRNN (32)
Dense (64, activation = “relu”)	Dense (64, activation = “relu”)	Dense (32)
Dropout (0.2)	Dropout (0.2)	Dropout (0.2)
Dense (2, activation = “softmax”)	Dense (2, activation = “softmax”)	Dense (16)
—	—	Dense (2, activation = “softmax”)
Loss = “binary_crossentropy,” optimizer = “Adam,” epochs = 100