Chemical-induced disease relation extraction via attention-based distant supervision

BMC Bioinformatics

Table 4 Hyper-parameters for two models

Method	Hyper-parameter	Value
Attention-based Model	Learning rate	0.004
	LSTM hidden state dimension	200
	Mini-batch size	500
	Word embedding dimension	300
	Position embedding dimension	50
	Identifier embedding dimension	100
	Hyponym embedding dimension	50
	Location embedding dimension	50
	Hidden layer nodes	250
	Dropout rate	0.3
Stacked Auto-encoder Model	Learning rate	0.008
	Mini-batch size	400
	Word embedding dimension	300
	Identifier embedding dimension	100
	Hyponym embedding dimension	50
	Encoder layer nodes	250
	Decoder layer nodes	50
	Dropout rate	0.3

ISSN: 1471-2105