Short-Term Daily Univariate Streamflow Forecasting Using Deep Learning Models

Table 2

Keras tuner optimized hyperparameter values for Borkena station with its MSE score.

Hyperparameters	MLP				S-LSTM				Bi-LSTM				GRU
	T + 1	T + 2	T + 3	T + 4	T + 1	T + 2	T + 3	T +	T + 1	T + 2	T + 3	T + 4	T + 1	T + 2	T + 3	T + 4
Layer_1_units	35	10	40	35	25	35	40	40	35	40	25	20	25	30	15	10
Dropout_1	0.0	0.0	0.0	0.1	0.1	0.0	0.2	0.2	0.1	0.2	0.0	0.1	0.1	0.0	0.1	0.1
Layer_2_units	10	20	25	15	35	25	30	10	20	30	20	30	20	25	40	25
Learning rate	0.0001	0.0001	0.0001	0.001	0.0001	0.0001	0.001	0.001	0.0001	0.001	0.001	0.001	0.0001	0.001	0.0001	0.001
Number of epochs	90	100	100	40	60	80	20	30	70	20	50	70	50	60	40	50
Number of batch sizes	20	20	30	30	30	40	20	10	40	20	20	30	50	30	10	90
Score (MSE)	0.099	0.099	0.096	0.097	0.099	0.099	0.098	0.096	0.099	0.099	0.097	0.096	0.099	0.099	0.099	0.097