On the stochastic acceleration of sequences of random variables (Q1334844)

This paper considers transformations changing a sequence \((s_ n)\) of numbers converging to \(s\) into another sequence \((t_ n)\) such that \((t_ n)\) converges to \(s\) faster than \((s_ n)\); such transformations are called accelerating the convergence of the initial sequence. In particular, the new notion of stochastic acceleration is defined for sequences of random variables converging almost completely (a.c.), almost surely (a.s.) or in probability (p.). Numerous helpful examples are given in order to clarify the definitions. Theorems of existence for such transformations are proved in the case of finite summation processes as well as for Aitken's \(\Delta^ 2\) process. Very interesting work in theoretical probability with possible applications in numerical analysis and statistical estimation.