MATH F113

(Probability and Statistics)

Chandra Shekhar
Associate Professor

Department of Mathematics
BITS Pilani, Pilani Campus, Rajasthan 333 031
Email: chandrashekhar@pilani.bits-pilani.ac.in
Mobile: 9414492349
Chandra Shekhar MATH F113 (Probability and Statistics)
What have you covered?

In Lecture 18
Problems on Normal Distribution
Log Normal Distribution

Chandra Shekhar MATH F113 (Probability and Statistics)

Normal Probability rule and Chebyshev’s
inequality

Normal Probability Rule and

Chebyshev’s inequality
It is sometimes useful to have a quick
way of determining which values of
a random variables are common and
which are considered to be rare.

Chandra Shekhar MATH F113 (Probability and Statistics)

Normal Probability rule and Chebyshev’s
inequality (Cont...)

Let X be normally distributed with

parameters mean µ and standard de-
viation σ then
P [−σ < X − µ < σ] = 0.68
P [−2σ < X − µ < 2σ] = 0.95
P [−3σ < X − µ < 3σ] = 0.997

Chandra Shekhar MATH F113 (Probability and Statistics)

Normal Probability rule and Chebyshev’s
inequality (Cont...)

Chandra Shekhar MATH F113 (Probability and Statistics)

Normal Probability rule and Chebyshev’s
inequality (Cont...)

The rule will be used later to

obtain a quick estimate of the
standard deviation of a normally
distributed random variable.
A second rule of thumb that can
be used to gauge the rarity of
observed values of a random
variable is Chebyshev’s
inequality.
Chandra Shekhar MATH F113 (Probability and Statistics)
Normal Probability rule and Chebyshev’s
inequality (Cont...)

Chebyshev’s Inequality: Let X

be a random variable with finite vari-
ance, then for any positive number
k,
1
P [|X − µ| ≥ kσ] ≤ 2
k
1
P [|X − µ| < kσ] ≥ 1 − 2
k
where µ the mean and σ standard
deviation.
Chandra Shekhar MATH F113 (Probability and Statistics)
Normal Probability rule and Chebyshev’s
inequality (Cont...)

Put kσ =  then
σ2
P [|X − µ| ≥ ] ≤ 2


Chandra Shekhar MATH F113 (Probability and Statistics)

Normal Probability rule and Chebyshev’s
inequality (Cont...)

Problem: Show, by Chebysheff ’s

inequality, that in 2000 throws with
a coin, the probability that the num-
ber of heads lies between 900 and
1100 is at least 19/20.

Chandra Shekhar MATH F113 (Probability and Statistics)

Normal Probability rule and Chebyshev’s
inequality (Cont...)

Solution: Let X be the number of

heads (success) in throwing of a coin,
then X ∼ Bin(n, p), where n = 2000,
p = 21 E(X) = np = 1000, V ar(X) =
npq = 500. Using, Chebysheff ’s inequal-
ity we get,

Chandra Shekhar MATH F113 (Probability and Statistics)

Normal Probability rule and Chebyshev’s
inequality (Cont...)

1
P [|X − µ| < kσ] ≥ 1 − 2
k
h √ i 1
P |X − 1000| < 500k ≥ 1 − 2
k
19
P [900 < X < 1100] ≥
20

Chandra Shekhar MATH F113 (Probability and Statistics)

Normal Probability rule and Chebyshev’s
inequality (Cont...)

Problem: If X is the number scored

in a throw of a fair die, show that the
Chebychev’s inequality gives
P [|X − µ| > 2.5] < 0.47
where µ is the mean of X, while the
actual probability is zero.

Chandra Shekhar MATH F113 (Probability and Statistics)

Normal Probability rule and Chebyshev’s
inequality (Cont...)

Solution: Let X be the random

variable which takes the values 1, 2,
3, 4, 5, 6, each with probability 1/6.
Then E(X) = 7/2 and V ar(X) = 35/12 =
2.9167. Now,
σ 2 2.9167
P (|X − µ| > ) < 2 =
 (2.5)2

Chandra Shekhar MATH F113 (Probability and Statistics)

Normal Probability rule and Chebyshev’s
inequality (Cont...)

P (|X − 3.5| > 2.5)

=P (X lies outside the limit 3.5−2.5, 3.5+
2.5) = 0

Chandra Shekhar MATH F113 (Probability and Statistics)

Normal Probability rule and Chebyshev’s
inequality (Cont...)

Exercise 50/4.5/pp.147: For a

normal distribution
P [|X − µ| < 3σ] = 0.997
What value is associated to this prob-
ability via Chebyshev’s inequality?
Are the results consistent? Which
rule gives a stronger statement in case
of a normal variable?
Chandra Shekhar MATH F113 (Probability and Statistics)
Normal Probability rule and Chebyshev’s
inequality (Cont...)

Exercise A symmetric die is thrown

600 times. Find the lower bound for
the probability of getting 80 to 120
sixes.

Chandra Shekhar MATH F113 (Probability and Statistics)

Normal Probability rule and Chebyshev’s
inequality (Cont...)

Exercise: For geometric distribu-

tion
f (x) = 2−x; x = 1, 2, 3...
prove that Chebyshev’s inequality gives
1
P [|x − 2| ≤ 2] >
2
15
while the actual probability is 16

Chandra Shekhar MATH F113 (Probability and Statistics)

Normal Probability rule and Chebyshev’s
inequality (Cont...)

Exercise: For the discrete variate

with density
x: -1 0 1
f (x): 1/8 6/8 1/8
Evaluate
P [|X − µx| ≤ 2σx]
Compare this result with that ob-
tained on using Chebyshev’s inequal-
ity
Chandra Shekhar MATH F113 (Probability and Statistics)
Normal Probability rule and Chebyshev’s
inequality (Cont...)

Exercise: How many times should

we toss a balanced coin so that we
can assert with probability at least
0.99 that the proportion of heads oc-
curs between 0.45 and 0.55?

Chandra Shekhar MATH F113 (Probability and Statistics)

Normal Probability rule and Chebyshev’s
inequality (Cont...)

Solution: Let n be the number of

trials and X be number of heads in
n trials
 
1
X ∼ BD n, p =
2
Hence, µ = n2 and σ 2 = n4
Proportion of heads Xn is between 0.45
and 0.55
if and only if 0.45n < X < 0.55n
Chandra Shekhar MATH F113 (Probability and Statistics)
Normal Probability rule and Chebyshev’s
inequality (Cont...)

0.45n < X < 0.55n = |X − µ| < 0.05n

By Chebyshev’s inequality
1
P (|X − µ| < kσ) ≥ 1 − 2
√
k
So, kσ = k 2n = 0.05n. Hence, k =
√
0.1 n
Thus if we choose n such that
1 100
1− 2 =1− > 0.99
k n
Hence, n > 10000
Chandra Shekhar MATH F113 (Probability and Statistics)
Normal Probability rule and Chebyshev’s
inequality (Cont...)

This justifies the relative frequency

approach to the probability. The pro-
portion of successes (i.e. relative fre-
quency) can be brought as close to
the actual probability by increasing
the number of trials.

Chandra Shekhar MATH F113 (Probability and Statistics)

Normal Approximation (Cont...)

Exercise: If X is Gamma random

variable with α = 0.05 & β = 100, find
an upper bound on
P ([(X − 4)(X − 6)] ≥ 999)

Chandra Shekhar MATH F113 (Probability and Statistics)

Normal Probability rule and Chebyshev’s
inequality (Cont...)

Exercise: From a usual pack of 52

cards, cards are drawn randomly with
replacement till the red card appears.
If X denotes the number of card drawn,
using Chebyshev’s inequality, find a
lower bound for
P [|X − 2| < 2]

Chandra Shekhar MATH F113 (Probability and Statistics)

Normal Probability rule and Chebyshev’s
inequality (Cont...)

Exercise: In one out of 6 cases,

material for bulletproof vests fails to
meet puncture standards. If 405 spec-
imens are tested, what does Cheby-
shev’s theorem tell us about the prob-
ability of getting at most 30 or more
than 105 cases that do not meet punc-
ture standards?

Chandra Shekhar MATH F113 (Probability and Statistics)

Normal Probability rule and Chebyshev’s
inequality (Cont...)

Solution: Let X be number of bul-

letproof vests out of 405 which fail
to meet puncture standards. Hence,
 
1
X ∼ BD n = 405, p =
6
√
So, µ = np = 67.5 and σ = npq = 7.5
Take kσ = 37.5 or k = 5
Required Probability
P (X ≤ 30 or X > 105)
Chandra Shekhar MATH F113 (Probability and Statistics)
Normal Probability rule and Chebyshev’s
inequality (Cont...)

Required Probability
P (X ≤ 30 or X > 105)
≤ P (X ≤ 30 or X ≥ 105)
1
= P (|X − µ| ≥ kσ) ≤ 2 = 0.04
k

Chandra Shekhar MATH F113 (Probability and Statistics)

Normal Probability rule and Chebyshev’s
inequality (Cont...)

Exercise 44/4.4/pp.146 Assum-

ing that during seasons of normal rain-
fall the water level in feet at a par-
ticular lake follows a normal distri-
bution with mean of 1876 feet and
standard deviation of 6 inches.

Chandra Shekhar MATH F113 (Probability and Statistics)

Normal Probability rule and Chebyshev’s
inequality (Cont...)

(a) During such a season, would it

be unusual to observe a water
level of at most 1875 feet.
(b) Suppose that the water will
crest the spillway if the water
level exceeds 1878 feet. What is
the probability that this will
occur during a season of normal
rainfall?
Chandra Shekhar MATH F113 (Probability and Statistics)
Normal Probability rule and Chebyshev’s
inequality (Cont...)

Exercise: A computer firm intro-

duces a new home computer. Past
experience shows that the random
variable X, the time of peak demand
measured in months after its intro-
duction, follows a gamma distribu-
tion with variance 36

Chandra Shekhar MATH F113 (Probability and Statistics)

Normal Probability rule and Chebyshev’s
inequality (Cont...)

(a) If the expected value of X is 18

months, find α and β
(b) Find P [X ≤ 7.01], P [X ≥ 26],
P [13.7 ≤ X ≤ 31.5]

Chandra Shekhar MATH F113 (Probability and Statistics)