Fusion Deep Learning and Machine Learning for Heterogeneous Military Entity Recognition

Table 4

Model hyperparameters.

Parameters	Values
Batch size	64
Clip	0.5
Dropout rate	0.5
Learning rate	10^-5
Maximum sequence length	512
BiLSTM hidden layer dimension	256
Optimizer algorithm	SGD