Why is the k$-$nearest neighbors regression called a lazy learner?

a$.$ Because the k$-$nearest neighbors regression does not use training data to estimate model parameters or determine variable splitting rules, but instead directly uses the

average of the k most similar training set observations as its predicted value for a new observation.

b$.$ Because k$-$nearest neighbors does not have embedded feature selection.

c$.$ Because the k$-$nearest neighbors regression requires the standardization or normalization of the variables.

d$.$ Because the k$-$nearest neighbors method does not require the use of a test set.