Borel right process
In the mathematical theory of probability, a Borel right process, named after Émile Borel, is a particular kind of continuous-time random process.
Let be a locally compact, separable, metric space.
We denote by the Borel subsets of .
Let be the space of right continuous maps from to that have left limits in ,
and for each , denote by the coordinate map at ; for
each , is the value of at .
We denote the universal completion of by .
For each , let
:
\mathcal F_t = \sigma\left\{ X_s^{-1}(B) : s\in[0,t], B \in \mathcal E\right\},
:
\mathcal F_t^* = \sigma\left\{ X_s^{-1}(B) : s\in[0,t], B \in \mathcal E^*\right\},
and then, let
:
\mathcal F_\infty = \sigma\left\{ X_s^{-1}(B) : s\in[0,\infty), B \in \mathcal E\right\},
:
\mathcal F_\infty^* = \sigma\left\{ X_s^{-1}(B) : s\in[0,\infty), B \in \mathcal E^*\right\}.
For each Borel measurable function on , define, for each ,
:
U^\alpha f(x) = \mathbf E^x\left[ \int_0^\infty e^{-\alpha t} f(X_t)\, dt \right].
Since and the mapping given by is right continuous, we see that
for any uniformly continuous function , we have the mapping given by is right continuous.
Therefore, together with the monotone class theorem, for any universally measurable function , the mapping given by , is jointly measurable, that is, measurable, and subsequently, the mapping is also -measurable for all finite measures on and on .
Here,
is the completion of
with respect
to the product measure .
Thus, for any bounded universally measurable function on ,
the mapping is Lebeague measurable, and hence,
for each , one can define
:
U^\alpha f(x) = \int_0^\infty e^{-\alpha t}P_tf(x) dt.
There is enough joint measurability to check that
\} is a Markov resolvent on ,
which uniquely associated with the Markovian semigroup .
Consequently, one may apply Fubini's theorem to see that
:
U^\alpha f(x) = \mathbf E^x\left[ \int_0^\infty e^{-\alpha t} f(X_t) dt \right].
The following are the defining properties of Borel right processes:{{harvnb|Sharpe|1988|loc=Sect. 20}}
- Hypothesis Droite 1:
:For each probability measure on , there exists a probability measure on such that is a Markov process with initial measure and transition semigroup .
- Hypothesis Droite 2:
:Let be -excessive for the resolvent on . Then, for each probability measure on , a mapping given by is almost surely right continuous on .
Notes
References
- {{citation|last=Sharpe|first=Michael|title=General Theory of Markov Processes|year=1988|isbn=0126390606}}
{{DEFAULTSORT:Borel Right Process}}