Quantum Coupon Collector
Abstract
We study how efficiently a -element set can be learned from a uniform superposition of its elements. One can think of as the quantum version of a uniformly random sample over , as in the classical analysis of the ``coupon collector problem.'' We show that if is close to , then we can learn using asymptotically fewer quantum samples than random samples. In particular, if there are missing elements then copies of suffice, in contrast to the random samples needed by a classical coupon collector. On the other hand, if , then quantum samples are~necessary. More generally, we give tight bounds on the number of quantum samples needed for every and , and we give efficient quantum learning algorithms. We also give tight bounds in the model where we can additionally reflect through . Finally, we relate coupon collection to a known example separating proper and improper PAC learning that turns out to show no separation in the quantum case.
Cite
@article{arxiv.2002.07688,
title = {Quantum Coupon Collector},
author = {Srinivasan Arunachalam and Aleksandrs Belovs and Andrew M. Childs and Robin Kothari and Ansis Rosmanis and Ronald de Wolf},
journal= {arXiv preprint arXiv:2002.07688},
year = {2020}
}
Comments
17 pages LaTeX