Hyperparameter
H
Hyperparameter
Definition
A configuration variable set before the training process that controls the learning behavior, such as learning rate, batch size, number of epochs, and network architecture. Unlike model parameters, hyperparameters are not learned from data.