On the weak laws for arrays of random variables (Q2483867): Difference between revisions

The authors prove the convergence in probability of sequences of the form \(\sum^{v_n}_{i= u_n} (X_{ni}- c_{ni})/b_n\), where \(\{u_n, n\geq 1\}\) and \(\{v_n, n\geq 1\}\) are sequences of integers, \(\{X_{ni}, u_n\leq i\leq v_n, n\geq 1\}\) are not necessarily independent random variables, \(\{c_{ni}, u_n\leq i\leq v_n, n\geq 1\}\) are constants or conditional expectations, and \(\{b_n, n\geq 1\}\) are constants growing unboundedly as \(n\to\infty\). Cesàro-type conditions are supposed that ensure the desired weak laws of large numbers without assumming explicit moment conditions on the random variables \(X_{ni}\).