Answers are grounded in this paper and cite their sources.
Let F̂ₙ denote the empirical distribution function for a sample of n i.i.d. random variables with distribution function F. In 1956 Dvoretzky, Kiefer and Wolfowitz proved that P(√n ₓ(F̂ₙ(x) - F(x)) > λ) ≤ C exp(-2λ²), where C is some unspecified constant. We show that C can be taken as 1 (as conjectured by Birnbaum and McCarty in 1958), provided that exp(-2λ²) ≤ 1/2. In particular, the two-sided inequality P(√n ₓ|F̂ₙ(x) - F(x)| > λ) ≤ 2 exp(-2λ²) holds without any restriction on λ. In the one-sided as well as in the two-sided case, the constants cannot be further improved.
No takes yet. Share an insight, caveat, or question.
Pascal Massart (1990) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: