Poisson probability

Discussion in 'Math' started by boks, Feb 23, 2009.

  1. boks

    Thread Starter Active Member

    Oct 10, 2008
    218
    0
    The number of ships X that arrives at a port during a day has proven to be given by a Poisson distribution with E(X) = 2. Which number of ships has largest probability of arriving on a certain day?


    Attempt.

    <br />
p(x;2) = \frac{e^{-2} \cdot 2^x}{x!}<br />

    The formula yields

    p(0;2) = 0.135
    p(1;2) = 0.270
    p(2;2) = 0.270
    p(3;2) = 0.180

    and then values of p approaching zero. The numbers of ships most likely to arrive are 1 or 2.

    The port can handle at most 3 ships a day. Redundant ships are sent to another port.
    What's the expected number of ships that are handled each day?

    Expected number= 1*0.27 + 2*0.27 + 3* 0.18 = 1.35 (Correct: 1.782)

    Which capacity is necassary to be able to handle all incoming ships on a day with 90% certainty?

    No idea...
     
    Last edited: Feb 23, 2009
  2. Mark44

    Well-Known Member

    Nov 26, 2007
    626
    1
    You have three questions here by my count. Here are some suggestions for answering them.
    1. "Which number of ships has largest probability of arriving on a certain day?"
    For what value of x (not limited to integer values) is p(x, 2) the largest? To do this, you would need to differentiate p(x, 2) set p'(x, 2) to zero, and solve for x. A complication is that p(x, 2) has a factorial in the denominator, but you can replace x! with Gamma(x + 1).

    2. "The port can handle at most 3 ships a day. Redundant ships are sent to another port. What's the expected number of ships that are handled each day?"
    This might be a naive approach, but isn't the expected number of ships E(x)? For a Poisson distribution, the mean and variance = lambda = E(x), which you give as 2.

    3. "Which capacity is necassary to be able to handle all incoming ships on a day with 90% certainty?"
    I think what you want here is the value of x for which p(x, 2) <= .9, which you should be able to find from a table of values for Poisson distributions. The idea is that you want to find the value of x for which 90% of the area under the curve lies to the left of x, and 10% of the area lies to the right.

    Hope these suggestions are helpful.
    Mark
     
Loading...