Table 5. Optimized hyperparameters for the ADX prediction model.

Layer	Searching range	Optimal value
LSTM 1	32 ∼ 256	224
Dropout 1	0.1 ∼ 0.5	0.3
LSTM 2	32 ∼ 256	192
Dropout 2	0.1 ∼ 0.5	0.1
LSTM 3	32 ∼ 256	96
Dropout 3	0.1 ∼ 0.5	0.1
LSTM 4	32 ∼ 256	160
Dropout 4	0.1 ∼ 0.5	0.5