Table 3 Parameters of the proposed deep neural model.

From: Panoramic tongue imaging and deep convolutional machine learning model for diabetes diagnosis in humans

Parameters

Parametric values

Learning rate

0.2

Number of epochs

Until the network settles at convergence criterion

Convergence criterion

10–6

Loss function (mean square error)

\(\sum\limits_{t = 1}^{n} {\left( {y_{t} - O{}_{t}} \right)^{2} }\)

Learning rule

Gradient Descent learning rule

Optimizer

Ant colony optimizer

Activation function

Gaussian function

Number of deep radial basis layers

5

Batch size

40 samples

Back to article page