Let
- E be a Polish space and E be the Borel σ-algebra on E
- I⊆[0,∞) be closed under addition and 0∈I
Please consider the following result:
Let (κt:t∈I) be a Markovian semigroup on (E,E) ⇒ There is a measurable space (Ω,A) and a Markov process X with distributions (Px)x∈E such that Px[Xt∈B]=κt(x,B)for all x∈E,B∈E and t∈I. Conversely, given a Markov process X with distributions (Px)x∈E on a measurable space (Ω,A), a Markovian semigroup (κt:t∈I) is defined by (1).
It turns out that X in the first part of the statement can be constructed as the family of coordinate maps on (Ω,A)=(EI,E⊗I).
I've seen that many authors assume that Markov processes are such coordinate maps. Why can they do that?
The statement above doesn't state, that given (Ω,A) there is one unique Markov process, does it? However, the finite-dimensional distributions of X, i.e. Px[X∈⋅]∘π−1Jfor J⊆I with |J|<∞, where πJ:EI→EJ are the canonical projections, are uniquely determined by (1).
Maybe (Px)x∈E (not only the finite-dimensional distributions) are uniquely determined by (1), if I⊆N0 or I is at least almost countable or when E is almost countable.
So, why does the stated result allows us to think about X as being uniquely determined?
No comments:
Post a Comment