Quadratically Tight Relations for Randomized Query Complexity
Abstract
Let be a Boolean function. The certificate complexity is a complexity measure that is quadratically tight for the zero-error randomized query complexity : . In this paper we study a new complexity measure that we call expectational certificate complexity , which is also a quadratically tight bound on : . We prove that and show that there is a quadratic separation between the two, thus gives a tighter upper bound for . The measure is also related to the fractional certificate complexity as follows: . This also connects to an open question by Aaronson whether is a quadratically tight bound for , as is in fact a relaxation of . In the second part of the work, we upper bound the distributed query complexity for product distributions by the square of the query corruption bound () which improves upon a result of Harsha, Jain and Radhakrishnan [2015]. A similar statement for communication complexity is open.
Keywords
Cite
@article{arxiv.1708.00822,
title = {Quadratically Tight Relations for Randomized Query Complexity},
author = {Dmitry Gavinsky and Rahul Jain and Hartmut Klauck and Srijita Kundu and Troy Lee and Miklos Santha and Swagato Sanyal and Jevgenijs Vihrovs},
journal= {arXiv preprint arXiv:1708.00822},
year = {2017}
}
Comments
14 pages