Summary of Specific Discrete R.V.’s and their Distributions 1. Bernoulli – Setting: Probability experiment with two possible oucomes: Success (S) or Failure (F) – Random Variable:X = 1 if success; X = 0 otherwise. X ∼ Ber(p) – Sample Space: Im(X) = {0, 1} – PMF: ( px (1 − p)1−x if x ∈ {0, 1} pX (x; p) = 0 otherwise – CDF: 0 if x < 0 P bxc x B1,p (x) = p (1 − p)1−x if 0 ≤ x < 1 i=0 1 if x ≥ 1 – Mean and Variance: E[X] = p, V ar(X) = p(1 − p) 2. Binomial – Setting: Count the number of successes in a sequence of n iid Bernoulli trials. – Random Variable: X = number of successes in a sequence of n iid Bernoulli trials. X ∼ Bin(n, p) – Sample Space: Im(X) = {0, . . . , n} – PMF: ( n pX (x; n, p) = px (1 − p)1−x if x ∈ {0, . . . , n} 0 otherwise x – CDF: 0 if x < 0 P bxc n x Bn,p (x) = p (1 − p)1−x if 0 ≤ x < n i=0 x 1 if x ≥ n – Mean and Variance: E[X] = np, V ar(X) = np(1 − p) 3. Geometric – Setting: Perform iid Bernoulli experiments until a success occurs. – Random Variable: X = trial on which the first success occurs. X ∼ Geo(p) – Sample Sapce: Im(X) = {1, 2, 3, . . .} – PMF: pX (x; p) = ( (1 − p)x−1 p if x ∈ {1, . . . , } 0 otherwise. – CDF: ( 0 if x < 1 Geop (x) = 1 − (1 − p)bxc if x ≥ 1 – Mean and Variance: E[X] = p−1 , V ar(X) = 1−p p2 4. Poisson – Setting: Number of occurrences of a relatively rare event, where the probability of the event is constant across time intervals or spatial regions – Random Variable: X = number of events. X ∼ P o(λ). – Sample Space: Im(X) = {0, 1, 2, . . .} – PMF: ( pX (x; λ) = λx exp(−λ) x! if x ∈ {0, 1, . . . , } 0 otherwise. – CDF: ( 0 if x < 0 P oλ (x) = Pbxc λx exp(−λ) i=0 x! if x ≥ 0 – Mean and Variance: E[X] = λ, V ar(X) = λ Important Connection Between Bernoulli Random Variables and Binomial Random Variables • A Binomial r.v. can be represented as the sum of iid Bernoulli r.v.’s: – Let X1, . . . Xn be iid Bernoulli r.v.’s. For i = 1 . . . , n, ( 1 if Success Xi = 0 otherwise. P (Xi = 1) = p = 1 − P (Xi = 0) – Let X = X1 + X2 + . . . + Xn. – Then, X ∼ Bin(n, p). – Verify, using the representation of X as a sum of iid Bernoulli r.v.’s, that E[X] = np and V ar(X) = np(1 − p): ∗ By linearity of expectation, E[X] = E[X1] + E[X2] + . . . + E[Xn] = p + p + . . . + p = np ∗ By independence, V ar(X) = V ar(X1) + V ar(X2) + . . . + V ar(Xn) = p(1 − p) + p(1 − p) + . . . + p(1 − p) = np(1 − p) – Note: A Bernoulli r.v. is a Binomial r.v. with n = 1. Modified Geometric • The Geometric r.v. tells us the trial on which the first success occurs in a sequence of iid Bernoulli experiments. Alternatively, we could keep track of the number of failures before the first success. • Let Y = number of failures before the first success in a sequence of iid Bernoulli experiments, each with success probability p. We will say that Y has a “Modified Geometric” distribution. • If X ∼ Geo(p), then Y = X − 1. • What is Im(Y )? Im(Y ) = {0, 1, . . .}. • What is the pmf of Y ? P (Y = k) = P (X = k + 1) = (1 − p)k+1−1p = (1 − p)k p, for k = 0, 1, . . . • What is E[Y ]? What is V ar(Y )? 1 1−p −1= p p 1−p V ar(Y ) = V ar(X − 1) = V ar(X) = 2 p E[Y ] = E[X − 1] = E[X] − 1 =