Distribution PMF MGF Mean Variance Parameters

1 1
Pm aj t 1
Pm 1
Pm 2
Uniform (a1 , . . . , am ) p(x) = m, x ∈ {a1 , . . . , am } m j=1 e m j=1 aj m j=1 (aj − ā) ai ∈ R
x 1−x t
Bernoulli(p) p(x) = p (1 − p) , x = 0, 1 (1 − p) + pe p p(1 − p) p ∈ [0, 1]
 
n x
Binomial(n, p) p (1 − p)n−x , x = 0, . . . , n ((1 − p) + pet )n np np(1 − p) n ∈ N, p ∈ [0, 1]
x
x
p(x) = λx! e−λ , x = 0, 1, . . .


Poisson(λ) exp λ(et − 1) λ λ λ>0
 
x−1 r  t
pe
r
r(1−p)
Negative Binomial(r, p) p (1 − p)x−r , x = r, r + 1, . . . 1−(1−p)et
r
p p2 r ∈ N, p ∈ [0, 1]
r−1
(nx)(r−x
m
) rn r n m (n+m−r)
Hypergeometric(r, n, m) n+m no closed form n+m (n+m)2 (n+m−1) r, n, m ∈ N
( r )

Table 1: Common Discrete Distributions: PMF, MGF, Mean, Variance, Parameters

Distribution PDF CDF MGF Mean Variance Parameters

1 x−a ebt −eat a+b (b−a)2
Uniform (a, b) f (x) = b−a , a ≤x≤b F (x) = b−a , a≤x≤b t(b−a) 2 12 a<b
−λx −λx λ 1 1
Exponential (λ) f (x) = λe , x≥0 F (x) = 1 − e , x≥0 λ−t , t < λ λ λ2 λ>0
λα α−1 −λx λ α α α


Gamma (α, λ) f (x) = Γ(α) x e , x≥0 no simple closed form , t<λ α > 0, λ > 0
1

λ−t λ λ2
(x−µ)2

1
F (x) = Φ x−µ exp µt + 12 σ 2 t2
2


2
Normal (µ, σ ) f (x) = √
σ 2π
exp − 2σ2 σ µ σ µ ∈ R, σ > 0

Table 2: Common Continuous Distributions: PDF, CDF, MGF, Mean, Variance, Parameters

Distribution PDF MGF Mean Variance Parameters

Chi-square (n) f (x) = 1
2n/2 Γ(n/2)
xn/2−1 e−x/2 , x ≥ 0 (1 − 2t)−n/2 , t < 1/2 n 2n n ∈ N (degrees of freedom)
n+1

 −(n+1)/2
Γ 2
t(n) f (x) = √ 2 n
1 + xn no closed form 0 (n > 1) n
n−2 , (n > 2) n ∈ N (degrees of freedom)
nπ Γ 2
2n2 (m+n−2)
F (m, n) ratio of two chi-square r.v.’s: U/m
V /n , where U∼ χ2m , V ∼ χ2n no closed form n
n−2 , (n > 2) m(n−2)2 (n−4) , (n > 4) m, n ∈ N (degrees of freedom)

Table 3: Common Sampling Distributions: PDF, MGF, Mean, Variance, Parameters

 Distribution (params) E[X] E[X2 ] m1 , m2 Moment equations & MME

Bernoulli(p) p p m1 = X̄, m2 = X 2 = X̄ E[X] = m1 ⇒ p̂ = m1 .

m1
Binomial(n, p) (known n) np np(1 − p) + n2 p2 m1 = X̄, m2 = X 2 E[X] = m1 ⇒ np = m1 ⇒ p̂ = .
n
Poisson(λ) λ λ + λ2 m1 = X̄, m2 = X 2 E[X] = m1 ⇒ λ̂ = m1 .
1 2 1 1
Exponential(λ) (rate) m1 = X̄, m2 = X 2 E[X] = m1 ⇒ = m1 ⇒ λ̂ = .
λ λ2 λ m1
θ θ2 θ
Uniform(0, θ) m1 = X̄, m2 = X 2 = m1 ⇒ θ̂ = 2m1 .
2 3 2
2

Normal(µ, σ 2 ) µ µ2 + σ 2 m1 = X̄, m2 = X 2 µ̂ = m1 , σ̂ 2 = m2 − m21 .

α α(α + 1)
Gamma(α, λ) (shape α, rate λ) m1 = X̄, m2 = X 2 From E[X] = αλ = m1 , (X) = λα2 = m2 − m21 :
λ λ2
m21 m1
α̂ = 2 , λ̂ = .
m2 − m 1 m2 − m21
1 2−p 1 1
Geometric(p) (support 1, 2, . . .) m1 = X̄, m2 = X 2 = m1 ⇒ p̂ = .
p p2 p m1

1 Pn
Table 4: Method of Moments: population moments, sample moments, equations, and MME estimators. Here m1 = n i=1 Xi , m2 =
1 Pn 2 2 2
n i=1 Xi , and s = m2 − m1 .