Stochastic Processes: Random Walk Analysis
Translated by
ScribdTranslationsStochastic Processes: Random Walk Analysis
Translated by
ScribdTranslationsAutonomous University of Guadalajara Stochastic Processes
Solution Task 1
1. A particle performs a symmetric random walk on Z starting at zero. Find
the probability that the particle will be back at the origin in the sixth step.
61 5
( 0|
6=0X= 0= )( ) =
3 2616
2. A particle performs a symmetric random walk on Z starting at zero. What is the
probability that the particle returns to the origin for the first time on the sixth step?
Seaτ = min(n > 0 | X = 0), then the sought probability can be written as
( = 6| 0= 0)
For a trajectory to be at the origin in step 6, it had to have gone up 3 times and
1
downloaded 3 times. Like in this case = = 2, the probability of each of those
1
trajectories is, so it is only necessary to count how many trajectories meet that
26
they return for the first time to the origin in the sixth step.
Analyzing the trajectories, we see that only the 4 trajectories that are present
above meet the criteria. Therefore, the sought probability is given by
4 1
( = 6| 0= 0= ) =4
22
6
3. Show that the moment generating function of the variable of the random walk
simple about Z is
From this expression, find again the expectation and variance of .
Since = =1∑ where { } they are i.i.d. random variables with
=1,2,.., (+1= y)
( = -1 =) we have the following equalities:
∑ − −
( )= ( =1 )= (∏ )= ∏( ) =∏ ( + )=( + )
=1 1 =1
By deriving once with respect to the previous expression, we obtain:
( + − ) = ( + − ) −1 ( − − )
Evaluating the derivative at = 0 we obtain that
( )= ( + ) -1 ( − )= ( − )
Deriving again with respect to we obtain:
2
( + − )
2
= (n - 1 ( ) + − ) −2 ( − − )2+ ( + − ) -1 ( + − )
Evaluating the second derivative at = 0 we obtain that
2)
( = ( - 1 () + ) -2
( − )2 + ( + ) -1 ( + )= (n - 1( ) − ) + 2
Therefore,
2) 2(
( ) = ( − =) −2 ( − ) 2 − ( + −) 2 − 2( )2
2
= − ( − )+ = 1 (− −( ) ) = (1 - p + q 1 )(
2 +p-q )
= (2 )(2 ) = 4npq.
4. Probability of ruin for the second player. If we now consider the game from the point of
view of player B, then it is the same random walk only now the capital
initial is N - k and the probability of winning in each bet is q. Substitute these parameters in
the solution to the gambler's problem and check that the probability of ruin for player B,
denoted by − , es la que aparece abajo. Verifique además que + − = 1, that is to say,
the probability that eventually the game will end with the ruin of one of the players
it's one.
Replacing the value = − in the formula of forp = 1/2, we have to
−( − )
− = =
Substituting the value = − y = in the formula of forp ≠ 1/2we have to
( / ) − −( / ) ( / ) (( / ) − - 1 ) ( / ) − -1 ( / ) -1
− = = = =
1 - (/ ) ( / ) (( / ) − - 1 ) ( / ) − -1 ( / ) -1
So the result of the exercise is fulfilled. Let's see now that + − = 1
Casep = 1/2:
−
+ − = + = equals 1
Casep ≠ 1/2:
( /) −( )/ ( / ) − 1( /) −( ) / (− ) / + ( 1 )1− /
+ − = + = =
1 − (/ ) ( / ) -1 1 − (/ ) 1 - (/ )
=1
5. (5/100 extra points on the first partial) Demonstrate that the average duration of the game in
the problem of the player's ruin is always less than or equal to /2, that( is, for
) 2 each
= 1,2, … , , it is fulfilled that ≤ /2both
( in)the symmetric case (2 pts) and in the non
2
symmetric (3 pts).
Case symmetric =
= ( − )
Deriving with respect to we have to
= − + (− N)- 2k
Solving the equation = 0to find the point where reaches the maximum,
we have
N - 2k = 0 ⇒ k = N/2
Therefore, reaches its maximum ink = N/2that is to say
≤ ( )( − ) N/2( )2
2 2
Symmetrical house ≠