Skip to main content

Table 7 Parameter setting regarding proposed LSTM + CNN model

From: Detection and classification of social media-based extremist affiliations using sentiment analysis techniques

LSTM + CNN model layers

Parameters with values

Convolutional layer

Number of filters = 2, 4, 6, 8, 10, 12, 14, 16

Kernel size = 2 × 2

Padding = same

Pooling layer

Pooling size = 2 × 2

LSTM layer

Units = 100

Additional

Vocabulary size = 2000

Input vector size = 50

Embedding dimension = 128

Batch size = 8

Number of epochs = 7