Improved Kolmogorov inequalities for the binomial distribution (Q1185548): Difference between revisions

The aim of this paper is to improve well-known Kolmogorov-type inequalities for the binomial distribution. E.g. Theorem 1 presents the following two-sided maximal large deviation probability inequality: Let \(Y_ 1,Y_ 2,\dots\) be a sequence of i.i.d. Bernoulli random variables with mean \(p\in(0,1)\). Then for all \(n\geq 0\) and \(0\leq\varepsilon<\min(p,1-p)\): \[ P(\sup_{k\geq n}|\overline{Y}_ k-p|\geq\varepsilon)\leq 2 \exp(-2n\varepsilon^ 2- {4\over9}n\varepsilon^ 4), \] where \(\overline{Y}_ k={1\over k}\sum_{j=1}^ k Y_ j\) and \(\overline{Y}_ 0=0\).