Table 2 DCNN hyperparameters optimized using the AO algorithm.

From: Deep convolutional neural network based archimedes optimization algorithm for heart disease prediction based on secured IoT enabled health care monitoring system

Hyperparameter

Description

Optimized value

Learning rate

Controls the speed of the gradient descent algorithm

0.0001

Momentum

Controls the impact of the previous weight update

0.900

Number of epochs

Determines how many times the training dataset parameters will be updated

30

Regularization

Overcomes the overfitting issue

0.2

Back to article page