|
|
vysotin@xxxxxxxxx wrote:
> Can someone help me in finding an expected time for standard linear
> random walk with two exit states to reach one exit state. Say, a walk
> with +1 and -1 steps (probabilities p and 1-p) starts at the position
> 0 and has final states at -k and +n. What would be the expected time to
> reach n?
The question has no meaning, since the walk might never reach n. You
can speak of the expected time to reach either -k or +n, or the
probability of reaching n, or the expected time to reach n _given that
n is reached_, etc. But the random variable T = time to reach n does
not exist in any reasonable sense, although I guess you could put T = +
infinity if n is never reached; this would give E(Time to reach n) =
infinity, too.
R.G. Vickson
> I am going to go through stopping time theory but for now I'd like to
> understand, if possible, in simple terms how such problems can be
> handled.
|
|