On computable learning of continuous features

Nathanael Ackerman, Julian Asilis, Jieqi Di +2

We introduce definitions of computable PAC learning for binary classification over computable metric spaces. We provide sufficient conditions for learners that are empirical risk m…