What is a hyperparameter in the context of machine learning?

In the context of machine learning, a hyperparameter is defined as a parameter that is set before the learning process begins. Hyperparameters are crucial because they influence the behavior and performance of the learning algorithm. They are not learned or adjusted from the training data but are instead predefined based on the model or based on empirical testing.

For example, in a neural network, hyperparameters could include the learning rate, the number of layers, the number of units in each layer, and the batch size. Choosing the right hyperparameters can significantly affect the model's ability to learn and generalize from the training data.

Other options describe elements that do not fit the definition of hyperparameters. A value learned during the training process refers to model parameters, which are adjusted based on the data. Data preprocessing techniques relate to methods applied to prepare data before training but are not hyperparameters themselves. A measure of model performance pertains to metrics used after the model has been trained to evaluate its effectiveness, which is distinct from hyperparameters.