Table 7 Summary of NN model architecture.

From: Predicting and classifying type 2 diabetes using a transparent ensemble model combining random forest, k-nearest neighbor, and neural networks

Hyperparameter

Search space

Hidden1

int [8, 128]

Hidden2

int [8, 256]

Hidden3

int [32, 256]

Dropout1

float [0.0, 0.5]

Dropout2

float [0.0, 0.5]

Dropout3

float [0.0, 0.5]

Learning_rate

log-uniform [1e−4, 1e−2]

Batch_size

categorical [16, 32, 64]

Epochs

int [50, 100]

Back to article page