Table 3.

The parameters in LSTM models used in this study. Time step refers to the length of input features used to make predictions. Loss function measures the difference between predicted and observed values. Number of units refers to the number of units in the LSTM layer. Epoch represents the number of completed training using all data in a training set. Batch size refers to the size of the input data used to update LSTM parameters one time. Learning rate refers to the rate for updating LSTM parameters. Optimizer refers to the algorithm for updating parameters. Dropout rate is the percent of units in the LSTM layer that is randomly discarded in the model training. The two groups of LSTM parameters were fixed separately by comparing the RMSE and MAE computed, based on validation, dataset.

Parameters	LSTM with NDVImean, RHmean, Rsum and Tmean	LSTM with Historical Dengue Data, NDVImean, RHmean, Rsum and Tmean
Time step	12	12
Loss function	MSE	MSE
Number of units	64	64
Epoch	1150	2000
Batch size	12	12
Learning rate	0.005	0.001
Optimizer	Adam	Adam
Dropout rate	0.8	0.65