Table 3

Experimental results of the BiLSTM model for different hyperparameters: learning rate, dropout, and gradient clipping norm

Learning rateDropoutClip normAccuracy
0.0010.330.330.69
0.00010.330.330.49
0.0010.50.50.72
0.0010.650.650.71

or Create an Account

Close subscription notice
Close access options