Scribd Logo
Upload

Welcome to Scribd!

  • 10 views

    Con Vol

    Uploaded by

    Diego Moreno
    The summary is: [1] The distribution of the sum of two independent random variables X and Y is given by the discrete convolution formula, which calculates the probability of their sum Z being any value z as the sum of the joint probabilities of X and Y summing to z. [2] The sum of independent binomial random variables has a binomial distribution with the sample sizes summed and the same probability of success. [3] Similarly, the sum of independent Poisson random variables has a Poisson distribution with the rate parameters summed.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as PDF, TXT or read online from Scribd

    Con Vol

    Uploaded by

    Diego Moreno
    10 views3 pages
    The summary is: [1] The distribution of the sum of two independent random variables X and Y is given by the discrete convolution formula, which calculates the probability of their sum Z being any value z as the sum of the joint probabilities of X and Y summing to z. [2] The sum of independent binomial random variables has a binomial distribution with the sample sizes summed and the same probability of success. [3] Similarly, the sum of independent Poisson random variables has a Poisson distribution with the rate parameters summed.

    Copyright

    © Attribution Non-Commercial (BY-NC)

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    The summary is: [1] The distribution of the sum of two independent random variables X and Y is given by the discrete convolution formula, which calculates the probability of their sum Z being any value z as the sum of the joint probabilities of X and Y summing to z. [2] The sum of independent binomial random variables has a binomial distribution with the sample sizes summed and the same probability of success. [3] Similarly, the sum of independent Poisson random variables has a Poisson distribution with the rate parameters summed.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    10 views3 pages

    Con Vol

    Uploaded by

    Diego Moreno
    The summary is: [1] The distribution of the sum of two independent random variables X and Y is given by the discrete convolution formula, which calculates the probability of their sum Z being any value z as the sum of the joint probabilities of X and Y summing to z. [2] The sum of independent binomial random variables has a binomial distribution with the sample sizes summed and the same probability of success. [3] Similarly, the sum of independent Poisson random variables has a Poisson distribution with the rate parameters summed.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    You are on page 1of 3

    Sums of Independent Random Variables

    Consider the sum of two independent discrete random variables X and Y whose values are restricted to the non-negative integers. Let fX () denote the probability distribution of X and fY () denote the probability distribution of Y . The distribution of their sum Z = X + Y is given by the discrete convolution formula. Theorem Discrete Convolution Formula. The random variable Z = X + Y has probability distribution fZ () given by
    z

    fZ (z) = fX+Y (z) = P ( Z = z ) =


    x=0

    fX (x)fY (z x)

    for z = 0, 1, ... . Proof: For each z, the event [Z = z] is the union of the disjoint events [X = x and Y = z x] for x = 0, 1, ..., z. Consequently, P ( Z = z ) = fZ (z) = =
    z x=0 z x=0

    P (X = x and Y = z x) fX (x)fY (z x)

    where the last step follows by independence. Let X1 and X2 be independent binomial random variables having the same probability of success. Their sum is again binomial. Corollary 1 Sum of Binomial Random Variables. Let X1 and X2 be independent binomial random variables where Xi has a Binomial(ni , p) distribution for i = 1, 2. Then

    X1 + X2

    has a binomial distribution with n1 +n2 trials and probability of success p

    Let X1 , X2 , ...Xk be independent binomial random variables where Xi has a Binomial(ni , p) distribution for i = 1, 2, ..., k. Then X1 + X2 + Xk has a Binomial(n1 + n2 + + nk , p) distribution.

    Proof: By the discrete convolution formula, Z = X1 + X2 has probability distribution 1

    P (X1 + X2 = z) = fZ (z) =
    x=0

    fX1 (x)fX2 (z x)

    so
    z

    fZ (z) =
    x=0

    n1 x n2 )p (1 p)n1 x ( )pzx (1 p)n2 (zx) x zx


    z

    = pz (1 p)n1 +n2 z
    x=0 y

    n1 n2 )( ) x zx

    Now, equating the coecients of s in the binomial expansion of both sides of (1 + s)n (1 + s)n = (1 + s)n1 +n+2 1 2 we conclude that
    z

    (
    x=0

    n + n2 n2 n1 ) ) = ( 1 )( z zx x

    The case for several binomial random variables follows by induction. Remark: Note that the sample sizes add but the success probability remains the same. Corollary 2 Sum of Poisson Random Variables. Let X1 and X2 be independent Poisson random variables where Xi has a Poisson (i ) distribution for i = 1, 2. Then X1 + X2 has a Poisson distribution with 1 + 2

    Let X1 , X2 , ...Xk be independent Poisson random variables where Xi has a Poisson (i ) distribution for i = 1, 2, ..., k. Then X1 + X2 + Xk has a Poisson(1 + 2 + + k ) distribution.

    Proof: By the discrete convolution formula, Z = X1 + X2 has probability distribution


    z

    P (X1 + X2 = z) = fZ (z) =
    x=0

    fX1 (x)fX2 (z x)

    so fZ (z) =

    z x=0

    x 1 zx 2 2 1 e e x! (z x)!
    z x=0

    = e(1 + 2 ) Use the binomial formula


    m

    x zx 2 1 x! (z x)!

    (a + b)m =
    x=0

    m x mx )a b x

    with m = z, a = 1 and b = 2 after multiplying and dividing by z!, to conclude that z x zx ( 1 + 2 )z 2 1 = z! x=0 x! (z x)! and the result is established. Remark: Note that the rate parameters i add.

    You might also like