Published 10 Jul 2020 in cs.DS and cs.DM | (2007.05634v2)
Abstract: A tantalizing conjecture in discrete mathematics is the one of Koml\'os, suggesting that for any vectors a<em>1,…,an∈B2<sup>m there exist signs x1,…,xn∈−1,1 so that ∣∑</em>i=1<sup>n</sup>xia<em>i∣</em>∞≤O(1). It is a natural extension to ask what ℓq-norm bound to expect for a<em>1,…,an∈Bp<sup>m. We prove that, for 2≤p≤q≤∞, such vectors admit fractional colorings x1,…,xn∈[−1,1] with a linear number of ±1 coordinates so that ∣∑</em>i=1<sup>n</sup>xia<em>i∣q≤O(min(p,log(2m/n)))⋅n<sup>1/2−1/p+</sup>1/q, and that one can obtain a full coloring at the expense of another factor of 1/2−1/p+1/q1. In particular, for p∈(2,3] we can indeed find signs x∈−1,1<sup>n with ∣∑</em>i=1<sup>n</sup>xia<em>i∣</em>∞≤O(n<sup>1/2−1/p</sup>⋅p−21). Our result generalizes Spencer's theorem, for which p=q=∞, and is tight for m=n. Additionally, we prove that for any fixed constant $\delta>0$, in a centrally symmetric body K⊆R<sup>n with measure at least e<sup>−δ</sup>n one can find such a fractional coloring in polynomial time. Previously this was known only for a small enough constant -- indeed in this regime classical nonconstructive arguments do not apply and partial colorings of the form x∈−1,0,1<sup>n do not necessarily exist.
“Emergent Mind helps me see which AI papers have caught fire online.”
Philip
Creator, AI Explained on YouTube
Sign up for free to explore the frontiers of research
Discover trending papers, chat with arXiv, and track the latest research shaping the future of science and technology.Discover trending papers, chat with arXiv, and more.