scientific article; zbMATH DE number 1340221

From MaRDI portal
Publication:4265760





Suppose that \({\mathbf N}= \{1,2,\dots\}\) is partitioned into \(p\) infinite subsets \({\mathbf N}_1,\dots,{\mathbf N}_p\), and let \(Y_1,\dots, Y_p\) be \(p\) random variables. Let \(\{X_n\}\) be a sequence of independent random variables such that \(X_n\) has the same distribution as \(Y_i\) whenever \(n\in{\mathbf N}\), \(1\leq i\leq p\). The reviewer [J. Math. Anal. Appl. 187, No. 2, 371-383 (1994; Zbl 0869.60064)] proved that \[ \sum^\infty_{n=1} P(|X_1+\cdots+ X_n|\geq \varepsilon n)<\infty,\quad \varepsilon >0,\tag{1} \] always implies \[ \sum^\infty_{n=1} \sum^n_{k=1} P(|X_k|\geq n)< \infty\tag{2} \] and \[ \lim_{n\to\infty}{1\over n} \sum^n_{k=1} E[X_k1_{\{|X_k|< n\}}]= 0,\tag{3} \] while, conversely, if (2) and (3) hold, and other two auxiliary conditions are fulfilled, then (1) holds. He also conjectured that without any extra conditions (1) holds if and only if (2) and (3) hold. The conjecture was independently disproved by \textit{A. R. Pruss} [ibid. 199, No. 2, 558-578 (1996; Zbl 0853.60042)] and the author [ibid. 206, No. 2, 606-610 (1997; Zbl 0878.60024)]. Moreover, Pruss provided some other weaker auxiliary conditions under which (2) and (3) imply (1). The purpose of this very well written paper is to improve on the works by the reviewer and Pruss. The main result is as follows: Assume there is a constant \(C_i\) such that \(\alpha_i(2n)\leq C_i\alpha_i(n)\), \(n\geq 1\),where \(\alpha_i(n)= \#\{l\in{\mathbf N}_i: l\leq n\}\), \(1\leq i\leq p\). Then (1) holds if and only if (2) and (3) hold.











This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4265760)