Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Limiting Distribution

  1. Oct 20, 2011 #1
    1. The problem statement, all variables and given/known data
    Suppose P(Xn = i) = [itex]\frac{n+i}{3n+6}[/itex], for i=1,2,3.
    Find the limiting distribution of Xn


    2. Relevant equations



    3. The attempt at a solution
    I first found the MGF by Expectation(etx)
    which resulted in etx([itex]\frac{n+1}{3n+6}[/itex] + [itex]\frac{n+2}{3n+6}[/itex] + [itex]\frac{n+3}{3n+6}[/itex])
    I then took the limit as n[itex]\rightarrow[/itex] ∞ which gives me 2etx

    Did I do this problem correctly? Is that the limiting distribution of Xn?

    Thanks.
     
  2. jcsd
  3. Oct 20, 2011 #2

    Ray Vickson

    User Avatar
    Science Advisor
    Homework Helper

    1) What you wrote is not the MGF.
    2) You don't need the MGF; in fact, using the MGF is doing it the hard way.

    RGV
     
  4. Oct 20, 2011 #3
    Do I have to calculate the CDF?
    After calculating the CDF and doing the limit as n goes to infinity, I get 1.

    Did I calculate the MGF incorrectly? Should I have done E(e^ti) instead?
     
  5. Oct 21, 2011 #4

    Ray Vickson

    User Avatar
    Science Advisor
    Homework Helper

    No, you did not calculate the MGF; you calculated exp(tx) times the sum of the probabilities-- in other words, just exp(tx), and for some completely undefined x. Yes, you should have calculated E(exp(t*i)), because that is exactly what the MGF is in this case (except you have used the symbol 'i' instead of 'X'). However, my original statement stands: you don't need the MGF, although using it correctly will do no harm.

    RGV
     
  6. Oct 21, 2011 #5
    No, you got it wrong. If you want to use the MGF method to find the limiting distribution, then, first, find the MGF of Xn for each n, which is

    [itex]MGF_{X_n}(t)=E[e^{tX_n}]=\sum_{k=1}^3 e^{tk}P(X_n=k)[/itex].

    Then, find the limit of the MGF as n tends to infinity. Finally, find some random variable X with MGF equal to [itex]\lim_{n\rightarrow\infty}MGF_{X_n}(t)[/itex].

    My suggestion is, instead using the MGF method, find the cumulative probability function, Fn(i)=P(Xn<=i) for each i=1,2,3. Then, limit this function as n tends to infinity. The limit is the cumulative probability function of the limiting distribution.
     
    Last edited: Oct 21, 2011
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?
Similar Discussions: Limiting Distribution
Loading...