Continuous

Probability
Models

Kaustav Banerjee

IIM Lucknow

July, 2022

K Banerjee (IIML) Continuous Probability Models July, 2022 1 / 21

Probability models for an ignorant

A simulator selects one of the digits (from 0 to 9) randomly. What is

the probability of getting 0?
A gambler has no reason to believe that the chosen dice for the game
is loaded. Which probability model is suitable?
A bank knows that the recovery on a loan, is anywhere between 60%
to 80% of the outstanding loan amount, in case of a default. In
absence of any other information, what is the probability that the
recovery rate will be between 60% to 70%? What is the probability
that the recovery rate will be between 70% to 80%?

K Banerjee (IIML) Continuous Probability Models July, 2022 2 / 21

Discrete uniform probability model

0.22
0.20
0.18
Probability

0.16
0.14
0.12
0.10

1 2 3 4 5 6

Outcomes

K Banerjee (IIML) Continuous Probability Models July, 2022 3 / 21

Continuous uniform probability model

0.07
0.06
Probability density

0.05
0.04
0.03

60 65 70 75 80

Rate

K Banerjee (IIML) Continuous Probability Models July, 2022 4 / 21

Uniform probability model
A random variable X has discrete uniform distribution over 1, 2, ..., N
with PMF
1
f(x) = , x = 1, 2, ..., N
N
where parameter N is a specified integer.
1. E(X) = (N + 1)/2, V(X) = (N2 − 1)/12
A random variable X has continuous uniform distribution over (α, β)
with PDF
1
f(x) = , α<x<β
β−α
where α, β ∈ < are two parameters.
1. E(X) = (α + β)/2, V(X) = (β − α)2 /12

K Banerjee (IIML) Continuous Probability Models July, 2022 5 / 21

Applications
A job takes anywhere between 0 and 1 hour to complete. (a) What
is the probability that it will take at most 5 minutes to complete?
(b) What is the probability that it will take 10 to 20 minutes to
complete? (c) What is the probability that it will take at least 45
minutes to complete if it has taken 30 minutes already?
Z x
1 x
P(X < x) = F(x) = dz =
0 β −α β−α
b−a
⇒ P(a < X < b) = P(X < b) − P(X < a) =
β−α

P(a < X < b) can be easily seen as an area of a rectangle, having

breadth b − a and length (probability density function) 1/(β − α)

K Banerjee (IIML) Continuous Probability Models July, 2022 6 / 21

Exponential probability model
A random variable X has exponential distribution with PDF
1 −x/λ
f(x) = e , 0<x<∞
λ
where λ > 0 is the parameter.
1. E(X) = λ, V(X) = λ2
2. For s > t, P(X > s|X > t) = P(X > s − t)
3. Like geometric, exponential is used to model ‘lifetime’ data
Jones figures the total number of thousands of miles an auto can be
driven before it would need to be junked is exponential with mean 20.
Smith has a used car that he claims has been driven only 10,000
miles. If Jones purchases the car, what is the probability that she
would get at least 20,000 additional miles out of it?

K Banerjee (IIML) Continuous Probability Models July, 2022 7 / 21

Exponential probability model

0.04
Probability density

0.02
0.00

0 20 40 60 80 100

Lifetime

K Banerjee (IIML) Continuous Probability Models July, 2022 8 / 21

Galton board game

Balls roll one-by-one down a board, striking the nails head-on

and deflecting to left/right, collected in the bins at the bottom
Each deflection can be coded by a random variable taking values
1(0) according as left (right) deflection
Regardless of the deflections in the previous rows, the ball is
equally likely to go left/right in current and future rows
Deflections in different rows are independent
Sum of the deflections indicate the position of the ball at the end
Sn = sum of the deflections with a board having n rows
Sn ∼ Binomial(n, 1/2) → true/false?
What will be the shape of the distribution of Sn for large n?

K Banerjee (IIML) Continuous Probability Models July, 2022 9 / 21

 0.25
0.20
0.15
Density

0.10
0.05
0.00

0 2 4 6 8 10

Sum.nrow

Figure: Galton board-game: 10 rows

K Banerjee (IIML) Continuous Probability Models July, 2022 10 / 21

 0.15
0.10
Density

0.05
0.00

10 15 20

Sum.nrow

Figure: Galton board-game: 30 rows

K Banerjee (IIML) Continuous Probability Models July, 2022 11 / 21

 0.10
0.08
0.06
Density

0.04
0.02
0.00

15 20 25 30 35 40

Sum.nrow

Figure: Galton board-game: 50 rows

K Banerjee (IIML) Continuous Probability Models July, 2022 12 / 21

 0.08
0.06
Density

0.04
0.02
0.00

35 40 45 50 55 60 65

Sum.nrow

Figure: Galton board-game: 100 rows

K Banerjee (IIML) Continuous Probability Models July, 2022 13 / 21

Approximation to binomial

Chance that the balls will be stored into the 4 middlemost bins, when
probability of moving into the right-sided bins is θ
θ = 0.5 θ = 0.25
n Binomial Approximation Binomial Approximation
10 0.7734375 0.7940968 0.2237091 0.3569928
30 0.5268703 0.5347912 0.0081291 0.0101665
50 0.4241507 0.4283924 0.0003633 0.0003015
100 0.3091736 0.3108435 0.0000002 0.0000000
500 0.1418094 0.1419723 0 0
1000 0.1005986 0.1006568 0 0

K Banerjee (IIML) Continuous Probability Models July, 2022 14 / 21

Normal distribution
A continuous random variable X has normal distribution with mean µ
and standard deviation σ, with PDF
(x − µ)2
1 −
f(x) = √ e 2σ 2 , −∞ < x < ∞
2πσ 2
where −∞ < µ < ∞ and σ > 0 are parameters.
X−µ
1. E(X) = µ, V(X) = σ 2 , X ∼ N(µ, σ 2 ) ⇒ ∼ N(0, 1)
σ
2. If X ∼ N(µ, σ 2 ), a and b are constants, a + bX ∼ N(a + bµ, b2 σ 2 )
3. If X1 ∼ N(µ1 , σ12 ) ⊥ X2 ∼ N(µ2 , σ22 ), and constants a1 , a2 ∈ <,
a1 X1 + a2 X2 ∼ N(a1 µ1 + a2 µ2 , a21 σ12 + a22 σ22 )

K Banerjee (IIML) Continuous Probability Models July, 2022 15 / 21

 Figure: N(µ = 0, σ 2 = 1)

Probability density

−5 −2 0 2 5

K Banerjee (IIML) Continuous Probability Models July, 2022 16 / 21

 Figure: N(µ = 0, σ 2 = 1); N(µ = 0, σ 2 = 9)

Probability density

−10 0 10

K Banerjee (IIML) Continuous Probability Models July, 2022 17 / 21

Computing normal probabilities
1. X ∼ N(µ = 3, σ 2 = 4). Compute the following:
P(X < 2)
P(X < −3)
P(−3 < X < 2)
97.5 percentile of X, i.e. k such that P(X < k) = 0.975
95 percentile of X, i.e. k such that P(X < k) = 0.95

2. GMAT scores for a group of students are approximately normally distributed

with mean 580 and SD = 55. Students above a score of 650 are admitted to a
business school.
What percentage of students are expected to be admitted to the school?
What percentage of the admitted students are expected to have a score
over 700?
What is the score of the student at the 95th percentile?

K Banerjee (IIML) Continuous Probability Models July, 2022 18 / 21

 Probability density

−3 −2 −1 0 1 2 3

P(|X − µ| < σ) = 0.683, P(|X − µ| < 2σ) = 0.954, P(|X − µ| < 3σ) = 0.997

K Banerjee (IIML) Continuous Probability Models July, 2022 19 / 21

Central Limit Theorem

Let X1 , X2 , ..., Xn be independent and identically distributed (IID)

random variables with mean E(Xi ) = µ and variance V(Xi ) = σ 2 .
Further, X̄ = 1n ni=1 Xi . Then, for large n,
P

n
X
Xi → N(nµ, nσ 2 ) as n → ∞
i=1
i.e. X̄ → N(µ, σ 2 /n) as n → ∞

1. CLT holds irrespective of the parent distribution for Xi

2. Normal is a ‘black hole’ among probability models
3. How large should n be, depends on the parent distribution of Xi

K Banerjee (IIML) Continuous Probability Models July, 2022 20 / 21

Galton board game

Xi : indicator taking 1(0) for the i’th deflection to right(left)

Xi ∼ Binomial(1, θ)
X1 , X2 , ..., Xn are independent and identically distributed
E(Xi ) = θ, V(Xi ) = θ(1 − θ)
Sn = ni=1 Xi indicates the position of a ball at the end bins
P

Sn ∼ Binomial(n, θ)
 p 
Sn → N nθ, nθ(1 − θ) as per CLT

K Banerjee (IIML) Continuous Probability Models July, 2022 21 / 21