Some strong limit theorems for the largest entries of sample correlation matrices (Q2494587): Difference between revisions

Let \(\{X_{k,i}: i \geq 1, k \geq 1\}\) be an array of independent and identically distributed random variables and let \(\{p_n, n \geq 1\}\) be a sequence of positive integers such that \(n/p_n\) is bounded away from zero and infinity. Let \(W_n= \max_{1 \leq i \leq j \leq p_n}| \sum_{k=1}^nX_{k,i}X_{k,j}| \) and \(L_n=\max_{1\leq i \leq j\leq p_n}| \hat \rho_{i,j}^{(n)}| \) where \(\hat \rho_{i,j}^{(n)}\) denotes the sample correlation coefficient between the vectors \((X_{1,i},\dots,X_{n,i})\) and \((X_{1,j},\dots,X_{n,j}).\) The authors derive strong limit theorems for the sequences \(W_n\) and \(L_n\). These results are obtained as corollaries for some general results proved for arrays of two-dimensional random vectors following the work of \textit{T. Jiang} [Ann. Appl. Probab. 14, No.~2, 865--880 (2004; Zbl 1047.60014)].