Which of the following statements about random forests is false?

a. The predictive performance of a random forest is often competitive with a boosting ensemble, and it is computationally much simpler.

b. A random forest is an ensemble consisting of a combination of different types of classifiers (k-nearest neighbors, logistic regression, neural networks, and trees).

c. A random forest is not as interpretable as an individual tree.

d. Correlation between individual trees in a random forest is reduced by limiting the number features considered at the variable splits.