Complexity of learning according to two models of a drifting environment

Abstract

We show that a cε3/VC dim(F) bound on the rate of drift of the distribution generating the examples is sufficient for agnostic learning to relative accuracy ε, where c>0 is a constant; this matches a known necessary condition to within a constant factor. We establish a cε2/VC dim (F) sufficient condition for the realizable case, also matching a known necessary condition to within a constant factor . We provide a relatively simple proof of a bound of O(1/ε2(VC dim (F)+log 1/δ)) on the sample complexity of agnostic learning in a fixed environment.