Skip to main content

Table 1 List of the hyperparameters optimized during the learning procedure

From: Biological interpretation of deep neural network for phenotype prediction based on gene expression

Hyperparameters Selected values Tested values
# hidden layers 3 1,2,3,4,5,6
# neurons 500/200/50 [20,1000]
\(p_{drop}\) 0.3 [0.1, 0.5]
\(\lambda \) \(10^{-3}\) \([10^{-2},10^{-5}]\)
lr \(10^{-4}\) \([10^{-2},10^{-5}]\)
batch size 16 [8, 256]