On the weak law of large numbers for normed weighted sums of i.i.d. random variables (Q803633): Difference between revisions

The main result of the paper is the following theorem: Let \(\{\) Y, \(Y_ n\), \(n\geq 1\}\) be i.i.d. random variables and let \(\{a_ n,\quad n\geq 1\},\{b_ n,\quad n\geq 1\}\) and \(c_ n=b_ n/| a_ n|\) be constants satisfying \(a_ n\neq 0\), \(b_ n>0\) and either \[ (i)\quad c_ n\nearrow,\quad c_ n/n\searrow,\sum^{n}_{j=1}a^ 2_ j=o(b^ 2_ n),\sum^{n}_{j=1}(c_ j/j)^ 2=O(b^ 2_ n/\sum^{n}_{j=1}a^ 2_ j), \] or \[ (ii)\quad c_ n/n\nearrow \text{ and } \sum^{n}_{j=1}a_ j^ 2=o(na^ 2_ n)\quad as\quad n\to \infty. \] If \(nP[| Y| >c_ n]=o(1)\), then \[ (\sum^{n}_{j=1}a_ j(Y_ j-E YI[| Y| \leq c_ n]))/b_ n\to^{P}_{n\to \infty}0. \] Also, related weak laws of large numbers for weighted random sums of i.i.d. random variables and weighted random sums normalized by random sequences are proved. The considerations are illustrated by suitable examples and counterexamples.