The Erdös discrepancy problem

From MaRDI portal



Abstract: We show that for any sequence taking values in −1,+1, the discrepancy sup_{n,d in {�f N}} left|sum_{j=1}^n f(jd) ight| of f is infinite. This answers a question of ErdH{o}s. In fact the argument also applies to sequences f taking values in the unit sphere of a real or complex Hilbert space. The argument uses three ingredients. The first is a Fourier-analytic reduction, obtained as part of the Polymath5 project on this problem, which reduces the problem to the case when f is replaced by a (stochastic) completely multiplicative function . The second is a logarithmically averaged version of the Elliott conjecture, established recently by the author, which effectively reduces to the case when usually pretends to be a modulated Dirichlet character. The final ingredient is (an extension of) a further argument obtained by the Polymath5 project which shows unbounded discrepancy in this case.


The main result in this paper is the following. Let \(H\) be a real or complex Hilbert space with norm \(\| \cdot \|_{H}\), and let \(f\) be a function defined on the positive integers \(\mathbb{N}\) such that \(\| f(n) \|_{H}=1\) for every \(n\in\mathbb{N}\). Then the discrepancy of \(f\), defined as NEWLINE\[CARRIAGE_RETURNNEWLINE\sup_{n,d\in\mathbb{N}}\left\| \sum_{j=1}^n f(jd) \right\|_{H},CARRIAGE_RETURNNEWLINE\]NEWLINE is infinite. As a special case, if \(f\) is a function taking values in \(\{-1,+1\}\), this result is the answer to an open question originally posed by Erdős, therefore known as the Erdős discrepancy problem.NEWLINENEWLINEThe proof of this theorem is based on a logarithmically averaged version of the Elliott conjecture, also recently shown by the author of this paper, and results obtained in the so-called \texttt{Polymath5} project.




Cited in
(42)








This page was built for publication: The Erdös discrepancy problem

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2826233)