Erdős Problem #992 — Metric discrepancy of dilated sequences
Let be any strictly increasing sequence of integers. For and positive integer , let denote the fractional part, and define the (unnormalized) interval discrepancy by
where is the length of . The question asks whether, for every such sequence and for Lebesgue-almost every , as ,
and, more strongly, whether one can always obtain
Erdős Problem #992 Aistleitner and Berkes, Probability and metric discrepancy theory
Background
Weyl’s metric theorem says that for any sequence of distinct integers , the fractional parts are uniformly distributed for almost every . Discrepancy measures the finite- error in this distribution. Erdős’s 1964 discussion asks how far the available almost-everywhere rate can be improved; the earlier metric bounds of Erdős–Koksma and Cassels provide its starting point. A standard analytic approach relates discrepancy to exponential sums involving through inequalities such as Erdős–Turán. Aistleitner and Berkes, Probability and metric discrepancy theory Erdős, Problems and results on diophantine approximations (1964)
References
UnsolvedMath, Erdős Problems set, ULAM AI, licensed CC BY 4.0.
Progress summary
Nothing recorded yet. Refresh searches the literature and the public web for attempts on this problem, and writes the first summary here.
Solutions 0
No solutions have been posted yet.