On the almost sure convergence for dependent random vectors in Hilbert spaces (Q2250828)

In this work two theorems on almost sure convergence of negatively associated random vectors in Hilbert spaces are developed. Extensions of a result by \textit{M.-H. Ko} et al. [J. Theor. Probab. 22, No. 2, 506--513 (2009; Zbl 1166.60021)] are given and illustrative examples are provided. The paper contains three parts: 1. Introduction. 2. Almost sure convergence: non-identically distributed cases, 3. The identically distributed case. The non-identically/identically distributed case means the case of sequences of non-identically/ identically distributed random vectors. The concept of negatively associated random variables was introduced by \textit{K. Joag-Dev} and \textit{F. Proschan} [Ann. Stat. 11, 286--295 (1983; Zbl 0508.62041)] and can be extended to random vectors. The strong convergence problem for sequences of negatively associated and/or positively associated random variables was investigated in other papers, e.g., by {B.-Y. Jing} and \textit{H.-Y. Liang} [J. Theor. Probab. 21, No. 4, 890--909 (2008; Zbl 1162.60008)], \textit{G. Xing} and \textit{S. Yang} [J. Theor. Probab. 23, No. 1, 169--192 (2010; Zbl 1192.60063)], and \textit{J.-F. Wang} and \textit{L.-X. Zhang} [Metrika 64, No. 3, 361--378 (2006; Zbl 1106.60036)].