Identifying health related occupations of Twitter users through word embedding and deep neural networks

BMC Bioinformatics

Table 1 Metrics scores

Model	Accuracy	Precision weighted	Recall weighted	F1 weighted score	Training time (s)
Naïve-Bayes	0.73	0.62	0.60	0.64	586
LR	0.78	0.70	0.70	0.70	574
SGD	0.79	0.74	0.78	0.75	607
GRU	0.89	0.83	0.86	0.88	357
LSTM	0.90	0.89	0.85	0.84	453
Bi-LSTM	0.91	0.86	0.85	0.88	762
BERT	0.94	0.90	0.89	0.89	908
ALBERT	0.95	0.92	0.91	0.90	350

ISSN: 1471-2105