hyperparameters

Source: Wikipedia

Parameters to control the process of learning the model, as opposed to being parameters of the model, which are to be learned.

Subsets:

• model hyperparameters (for model selection), e.g.
  • size of the NN
  • topology of the NN
• algorithm hyperparameters (affect the efficiency of the learning process), e.g.
  • learning rate
  • batch size

Source: google-ml-course

• Learning rate $\alpha$
• Regularisation rate $\lambda$
• Embedding dimensions