Superpolynomial Lower Bounds for Learning Monotone Classes
Abstract
Koch, Strassle, and Tan [SODA 2023], show that, under the randomized exponential time hypothesis, there is no distribution-free PAC-learning algorithm that runs in time for the classes of -variable size- DNF, size- Decision Tree, and -Junta by DNF (that returns a DNF hypothesis). Assuming a natural conjecture on the hardness of set cover, they give the lower bound . This matches the best known upper bound for -variable size- Decision Tree, and -Junta. In this paper, we give the same lower bounds for PAC-learning of -variable size- Monotone DNF, size- Monotone Decision Tree, and Monotone -Junta by~DNF. This solves the open problem proposed by Koch, Strassle, and Tan and subsumes the above results. The lower bound holds, even if the learner knows the distribution, can draw a sample according to the distribution in polynomial time, and can compute the target function on all the points of the support of the distribution in polynomial time.
Cite
@article{arxiv.2301.08486,
title = {Superpolynomial Lower Bounds for Learning Monotone Classes},
author = {Nader H. Bshouty},
journal= {arXiv preprint arXiv:2301.08486},
year = {2023}
}