Khinchin's Theorem On The Factorization Of Distributions
   HOME

TheInfoList



OR:

Khinchin's theorem on the factorization of distributions says that every
probability distribution In probability theory and statistics, a probability distribution is a Function (mathematics), function that gives the probabilities of occurrence of possible events for an Experiment (probability theory), experiment. It is a mathematical descri ...
''P'' admits (in the convolution semi-group of probability distributions) a factorization :P = P_1 \otimes P_2 where ''P''1 is a probability distribution without any indecomposable factor and ''P''2 is a distribution that is either degenerate or is representable as the convolution of a finite or countable set of indecomposable distributions. The factorization is not unique, in general. The theorem was proved by A. Ya. Khinchin for distributions on the line, and later it became clear that it is valid for distributions on considerably more general groups. A broad class (seeI.Z. Ruzsa, G.J. Székely, "Algebraic probability theory", Wiley (1988)) of topological semi-groups is known, including the convolution semi-group of distributions on the line, in which factorization theorems analogous to Khinchin's theorem are valid.


References

Theory of probability distributions {{mathanalysis-stub