Author response to reviews of " Decision trees as partitioning machines to

Neural Information Processing Systems 

About "VC bounds based on parametrized function classes": We do not think that these bounds are relevant