Which of the following best describes reinforcement learning?
a. The model determines how elements of the dataset are alike.
b. The training data match inputs to nodes in the network.
c. The training data contain input–output pairs.
d. The model learns by trial-and-error.
e. None of these