The strong law of large numbers says that, for real independent and identically distributed random variables with ,The assertion is almost sure convergence of the entire sequence of sample means. Finite first absolute moment suffices; a finite variance is not required.
Articles by others on the same topic
There are currently no matching articles.