Deep-Learning Prediction Model with Serial Two-Level Decomposition Based on Bayesian Optimization

Table 1

Bayesian optimization hyperparameter space and search results.

Hyperparameters	Type	Range of values	The optimized hyperparameter for	The optimized hyperparameter for
No. 1 hidden units	Integer	{48, 42, 36, 30, 24}	36	48
Batch size	Integer	{1, 2, 4, 8, 16, 32}	32	2
Optimizer	Categorical	{Adam, Nadam, Adadelta}	Nadam	Nadam