Approximating the Number of Relevant Variables in a Parity Implies Proper Learning

Open in new window