How can dimensionality reduction help address the curse of dimensionality?

A. It guarantees a larger training set B. It replaces numerical features with class labels C. It can remove irrelevant dimensions that make distances and density estimates unreliable D. It makes every pair of points equally distant