Two-parameter markov processes
Shou jun luo · Stochastics and stochastics reports · 1992
We combine one-parameter transition functions and the Cairoli's three-point transition functions to construct two~parameter regular Markov processes, which are (*)-Markov processes and Markov fields. Wide-past Markov processes with any form of “initial laws” on ∂R(2 ,+) are also constructed. But a counter example shows that even if such a process equals on ∂R(2,+) the same random variable, its one-parameter subprocesses may not be Markovian. Hence it is not a (*)-Markov process.