This article is cited in 11 papers

Short Communications

On stochastic approximation processes with infinite variance

T. P. Krasulina

Leningrad

Abstract: A stochastic approximation algorithm in the presence of noise is considered. The noise is assumed to have the moment of order $p<2$, its variance may be infinite.
Convergence of stochastic approximation process with probability 1 and in average is proved.

 English version: Theory of Probability and its Applications, 1969, 14:3, 522–526

Bibliographic databases:

• 
•