Skip to main content

Table 1 Value range of hyperparameter to be optimized

From: Forecasting VaR and ES by using deep quantile regression, GANs-based scenario generation, and heterogeneous market hypothesis

Hyperparameter

Symbol

Value

Determination method

Number of features

\(N_{f}\)

[1,5]

Trial and error approach

Number of hidden units

\(N_{h}\)

[5,50]

Bias-variance tradeoff

Mog step

\(r\)

[0,10]

Preset

Number of layers

\(N_{L}\)

[1,4]

Bias-variance tradeoff

Layer type

\(LT\)

[0,1]

Model selection principle

Epoch

\(epoch\)

[50,500]

Trial and error approach