PROBABILITY AND MOMENT INEQUALITIES UNDER DEPENDENCE - Rutgers.pdfVIP

Weidong Liu, Han Xiao and Wei Biao Wu Probability and moment inequalities play an important role in the study of properties of sums of random variables. A number of inequalities have been derived for independent random variables; see the recent collection by Lin and Bai (2010). The celebrated Nagaev and Rosenthal inequalities are two useful ones. We first start with the Nagaev inequality. Let X , . . . , X be mean 0 n independent random variables and S = ∑n X . Further assume that for all n i i p /p Nagaev (1979), for a i, ∥X ∥ := ( |X | ) ∞, p 2. By Corollary 1.7 in i p i positive number x, one has n { } ∑ a x p ∑ (S ≥ x) ≤ (X ≥ y ) + exp − n i i n (X X ≤y ) i i i (∑n (Xp ≤X ≤y ) )βx/y + i ip − , (1.1) βxy where y , . . . , y 0, y = max {y , 1 ≤ i ≤ n}, β = p/ (p + 2) and a = 2e−p (p + n i i p 2)− . With n