Table 2.

List of hyperparameters used for deep learning model training.

Hyperparameters	Value
Dimension of BioWordVec embeddings	200
Dimension of clinicalBERTa embeddings	768
BiLSTMb hidden size	256
Number of attention heads	2
Epochs	100
Patience	5
Iterations per epoch	100
Dropout rate	0.5
Learning rate	0.001
Batch size	32
Epochs for training BioWordVec embeddings	2

^aclinicalBERT: clinical bidirectional encoder representations from transformers.

^bBiLSTM: bidirectional long short-term memory.