Which of the following statements about the k-nearest neighbors classifier is true?

a. The decision boundary corresponding to a k-nearest neighbor classifier is a linear function of the input variables.

b. The decision boundary corresponding to a k-nearest neighbor classifier can be a highly nonlinear function of the input variables.

c. When applying k-nearest neighbors, one should include as many input variables as possible.

d. Standardizing the input variables is not necessary before applying a k-nearest neighbor classifier.