MOVES Seminar 9. November 2006

From Measure Theory to CTMDPs

 

Martin Neuhäusser

 

Abstract:

 

In this talk, I will give a short introduction to measure theory which will
be tailored to the definition of a semantics for Continuous Time Markov
Decision Processes.
Hopefully, the talk will also shed some light on the measure theoretic
concepts that are common in stochastic modelling, like the well-known
cylinder set construction.

In a second part, these concepts are used to define the semantics for CTMDPs.
At the end, some preliminary results regarding a logical characterisation
of CTMDP will be given.