Given a filtered probability space , a Markov process is an -adapted Stochastic Process with Markov Semigroup such that
for all measurable and bounded. As a result, the probability of being in state given the past is only a function of the process at time i.e. :