Show that the bayes classifier will achieve the best error rate, defined as:
$$ E(f) = \int \int \mathbb{I}(y = f(x)) \cdot p(x, y) dxdy $$
where $$f(x)$$ is the classifier, and $$p(x, y)$$ is the intrinsic data distribution.
I just want to get some help as to how to proceed.