Modeling and Analysis of Data-Driven Systems through Computational Neuroscience Wavelet-Deep Optimized Model for Nonlinear Multicomponent Data Forecasting

Table 2

Hyperparameters selected by Bayesian optimization.

Hyperparameters	Type	Bayesian optimization	Random search
Wavelet decomposition layer	Integer	8	10
Mother wavelet function	Categorical	db35	Sym7
No. 1 hidden units	Integer	48	24
Dropout rate	Uniform	0.0687	0
Batch-size	Integer	5	1
Epochs	Integer	350	300
Optimizer	Categorical	Adadelta	Adam