In this paper an irreducible parameterization for a finite memory sour
ce is constructed in the form of a tree machine. A universal informati
on source for the set of finite memory sources is constructed by a pre
dictive modification of an earlier-studied algorithm-Context. It is sh
own that this universal source incorporates any minimal data-generatin
g tree machine in an asymptotically optimal manner in the following se
nse: the negative logarithm of the probability it assigns to any long
typical sequence, generated by any tree machine, approaches that assig
ned by the tree machine at the best possible rate.