Discrete Probability Distributions

NOTE : This document could contain errors, please be warned. The text could possibly use other
notation and format. Also note that the PMF is P {X = x}, we call it p(x) but others call it f (x).

1. Discrete Uniform Distribution

Description : If a random variable X assumes the values x1 , x2 , . . ., xk , with equal

probabilities, then X is a discrete uniform random variable.

PMF : p(x) = k1 , x = x1 , x2 , . . . , xk .
k
X
xi
i=1
Mean : E[X] =
k
k
X
(xi E[X])2
i=1
Variance : V [X] =
k
2. Bernoulli Distribution

Description : A trial can result in a success with probability p and a failure with proba-
bility q (q = 1 p). Then the random variable X, which takes on 0 if the trial is a failure
and 1 if the trial is a success, is called the Bernoulli random variable with parameter p.

PMF : p(x) = px + (1 p)(1 x), x = 0, 1.

Mean : E[X] = p

Variance : V [X] = p(1 p)

3. Binomial Distribution

Description : A trial can result in a success with probability p and a failure with proba-
bility q (q = 1 p). Then the random variable X, the number of successes in n independent
trials is called the binomial random variable with parameters n and p.
n
x nx
PMF : p(x) = x p q , x = 0, 1, 2, . . . , n.

Mean : E[X] = np

Variance : V [X] = npq

1
4. Geometric Distribution

Description : A trial can result in a success with probability p and a failure with prob-
ability q (q = 1 p). Then the random variable X, denoting the number of trials until a
success is obtained is the geometric random variable with parameter p.

PMF :
p(x) = pq x1 , x = 1, 2, . . . , .

Mean : E[X] = 1/p

1p
Variance : V [X] = .
p2
5. Negative Binomial Distribution

Description : A trial can result in a success with probability p and a failure with prob-
ability q (q = 1 p). Then the random variable X, denoting the number of the trial on
which the k th success occurs is the negative binomial random variable with parameters k
and p.

PMF : !
x 1 k xk
p(x) = p q , x = k, k + 1, k + 2, . . . , .
k1

Mean : E[X] = k/p

k(1 p)
Variance : V [X] = .
p2
6. Hypergeometric Distribution

Description : A random sample of size n is selected without replacement from N

items. Of the N items, k may be classified as successes and N k are classified as failures.
The number of successes, X, in this random sample of size n is a hypergeometric random
variable with parameters N , n and k.

PMF :
k
N k

x nx
p(x) = N

, x = 0, 1, 2, . . . , n.
n

Mean : E[X] = nk/N

N n k k
 
Variance : V [X] = n 1 .
N 1 N N

2
7. Poisson Distribution

Description : A Poisson random variable X with parameter , represents the number of

outcomes occurring in a particular time interval.

PMF :
e ()x
p(x) = , x = 0, 1, 2, . . . ,
x!
where is the average number of outcomes in that time interval.

Mean : E[X] =

Variance : V [X] =

3
 Continuous Probability Distributions

NOTE : This document could contain errors. The text could possibly use other notation and format.
dF (x)
Note that the CDF, F (x) is P {X x}. Also the PDF, f (x) satisfies f (x) = .
dx

1. Exponential Distribution with parameter

PDF : (
ex , x>0
f (x) =
0 elsewhere
CDF : (
1 ex , x>0
F (x) =
0 elsewhere

1
Mean : E[X] =

1
Variance : V [X] =
2
2. Erlang Distribution with parameters and k

PDF : ( k1
(x)
(k1)! e
x x>0
f (x) =
0 elsewhere
CDF :
k1

(x)r
ex
X

1 x>0
F (x) = r!
r=0
0 elsewhere

Mean : E[X] = k/

Variance : V [X] = k/2

3. Gamma Distribution with parameters and

PDF : (
1 1 ex/
f (x) = () x x>0
0 elsewhere
Z
where () = x1 ex dx. If is an integer, then () = ( 1)!.
0
CDF : There is no closed form expression for the CDF. For numerical values, use tables
or software packages.

Mean : E[X] =

Variance : V [X] = 2

4
4. Continuous Uniform Distribution with parameters A and B

PDF : (
1
BA , AxB
f (x) =
0 elsewhere
CDF :
0
x<A
xA
F (x) = BA , AxB

1 x>B
A+B
Mean : E[X] =
2
(B A)2
Variance : V [X] =
12
5. Weibull Distribution with parameters and

PDF : (
x1 ex x>0
f (x) =
0 elsewhere
CDF : (
1 ex x>0
F (x) =
0 elsewhere
1
 
Mean : E[X] = 1/ 1 +

(  2 )
2 1
  
2/
Variance : V [X] = 1+ 1+

6. Normal Distribution with parameters and 2

PDF :
1 2
f (x) = e(1/2)[(x)/] < x <
2
CDF : There is no closed form expression for the CDF. For numerical values, use tables
or software packages.

Mean : E[X] =

Variance : V [X] = 2

7. Chi-Squared Distribution with parameter v

Special case of the Gamma distribution where = v/2 and = 2, where v is a positive
integer known as the degrees of freedom.

5
 PDF :
1
xv/21 ex/2
(
2v/2 (v/2)
x>0
f (x) =
0 elsewhere
Z
where (a) = xa1 ex dx.
0
CDF : There is no closed form expression for the CDF. For numerical values, use tables
or software packages.

Mean : E[X] = v

Variance : V [X] = 2v

8. Lognormal Distribution with parameters and 2

A continuous r.v. X has a lognormal distribution if the random variable Y = ln(X) has a
normal distribution with mean and standard deviation .

PDF :
2
(
1
2x
e(1/2)[(ln(x))/] x>0
f (x) =
0 elsewhere
CDF : There is no closed form expression for the CDF. For numerical values, use tables
or software packages.
2 /2
Mean : E[X] = e+
2 2
Variance : V [X] = e2+ (e 1)

9. Beta Distribution with parameters and

PDF : (
(+) 1
()() x (1 x)1 0<x<1
f (x) =
0 elsewhere
Z
where (a) = xa1 ex dx.
0
CDF : There is no closed form expression for the CDF. For numerical values, use tables
or software packages.

Mean : E[X] =
+

Variance : V [X] =
( + )2 ( + + 1)