The strong law of large numbers for weighted averages under dependence assumptions (Q1923941): Difference between revisions

The authors prove strong laws of large numbers for weighted averages of dependent random variables, generalizing the classical work of \textit{B. Jamison}, \textit{S. Orey} and \textit{W. Pruitt} [Z. Wahrscheinlichkeitstheorie Verw. Geb. 4, 40-44 (1965; Zbl 0141.16404)] for i.i.d. sequences. The dependence structure imposed is asymptotic quadrant sub-independent, requiring that \[ P(X_i>s, X_j>t) - P(X_i>s) P(X_j>t) \leq q \bigl(|i-j |\bigr) \alpha_{ij} (s,t), \] together with a similar condition on \(P(X_i<s, X_j<t)\). This condition generalizes the notion of asymptotic quadrant independence, introduced by \textit{T. Birkel} [Stat. Probab. Lett. 7, No. 1, 17-20 (1988; Zbl 0661.60048)]. The authors also prove a Marcinkiewicz-Zygmund SLLN for weighted averages. The proofs make heavy use of unpublished results by the same authors.