Upload
aruparna-maity
View
212
Download
0
Embed Size (px)
Citation preview
8/17/2019 Small Mathematical Expectation
1/6
1
Mathematical Expectation
of a Random Variable
2
Idea: mean value of a random
variable
• Definition: Weighted mean of the values ofthe random variable
• Condition for the existence:
– Note: Infinite sets can yield paradoxes
∑∈
== X x
x X xP X E )()(
∞
8/17/2019 Small Mathematical Expectation
2/6
7
Idea: mean value of a random
variable• Note that it is probab i l i s t i c in te rp re tat ion of the
common sense concept of mean .
• Note that if all possible samples are equallyprobable, they are equivalent.
∑∈= X x x N X Mean1
)(
1 3 0 1 4 0 1 5 0 1 6 0 1 7 0 1 8 0 1 9 0 2 0 0 2 1 0 2 2 00
5
10
15
20
25
1 3 0 1 4 0 1 5 0 1 6 0 1 7 0 1 8 0 1 9 0 2 0 0 2 1 0 2 2 0 2 3 00
10
20
30
40
50
60
70
80
90
1001 2 0 1 4 0 1 6 0 1 8 0 2 0 0 2 2 0 2 4 00
5
10
15
20
25
1 2 0 1 4 0 1 6 0 1 8 0 2 0 0 2 2 0 2 4 00
10
20
30
40
50
60
70
80
90
10 0
X X
∑∈ == X x x X xP X E )()( N x X P 1
)( ==
8
Idea: center of mass/equilibrium
• If the probability P(X =x) is interpreted asmass, and the random variable X asdistance, the mathematical expectation isthe center of mass of the object.
∑∈
== X x
x X xP X E )()(
9
Example: Roulette
• What is the payoff given by the casino?• Roulette selects at random a number between 1 and 36 (0) .
• Players can bet on 18,12,9,6,4,3,2,1 number.
– A Bet over k numbers has a probability of success of k/36 andof getting a payment x from the casino or lossing a.
– Expected value is
– A fair game would imply
( )ak
xk
x X xP X E LossWin x
−
−+=== ∑
∈ 361
36)()(
},{
ak
x
X E
−=
=
136
0)(
10
Example: Roulette
• What is the payoff given by the casino?
• The roulette has 37 slots (1 to 36+ the 0 slot) – In a Real game the casino the odds are
• 35:1 17:2 11:3
( )3737
137
136
)()(},{
ak a
k a
k x X xP X E
LossWin x
−=
−−+
−=== ∑
∈
0(
8/17/2019 Small Mathematical Expectation
3/6
13
Expected value of a geometric
random variable
• Random Varible X={Number of trials untila success}
• How do we sum this series?
L1,2,3,ifor)1()( 1 =−== − p pi X P i
( )L++++−=−= ∑
∞
=
− 32
0
1
4321)1()1()( p p p p pnp X E n
n
14
Expected value of a geometric
random variable
• How do we sum this series?
15
Expected value of a geometric
random variable
• How do we sum this series?
16
Expected value of a geometric
random variable
• How do we sum this series?
( )
( )
( )
1)1(
1)1()1(
1)1()()(
432)1()1()(
4321)1()1()(
Series
Geometric
0
432
432
1
32
0
1
=
−
−=−=
=+++++−=−
++++−=−=
++++−=−=
∑
∑
∑
∞
=
∞
=
∞
=
−
p p p p
p p p p p X pE X E
p p p p p pnp X pE
p p p p pnp X E
k
k
n
n
n
n
L
L
L
)1(
1 )(
p X E
−=
17
Expected value of a geometric
random variable• How do we sum this series? Another way
2
21
1
Series Geometric
1
)(
)()()()(
)(
)(
)1(
1
)1()(
)1()(
xb
xbdx
d xa xa
dx
d xb
xb
xa
dx
d
p p
p
dp
d kp pS
dp
d
p
p p pS
k
k
k
k
−=
−=
−
==
−==
∑
∑∞
=
−
∞
=
2)1(
1)1()(
p p X E
−−=
18
Expected value of a geometric
random variable• Random Varible X={Number of trials until
a success}
L1,2,3,ifor)1()( 1 =−== − p pi X P i
( )
p X E
p p p p pnp X E n
n
−=
++++−=−=∑∞
=
−
1
1)(
4321)1()1()(32
0
1L
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
1 6 1 1 1 6 2 1 2 6 3 1 3 6 4 1 4 6 5 1 5 6 6 1
9
1
1
1)( =
−=
p X E
8/17/2019 Small Mathematical Expectation
4/6
19
Relation of expectaition and other
statistical measures
• Skew distribution vs. symetric distribution
Full House: The Spread of Excellence from
Plato to Darwin by Stephen Jay Gould
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 50
1
2
3
4
-8 -6 -4 -2 0 2 4 6 80
0.2
0.4
0.6
0.8
Mode Median MeanMode Median Mean
20
Property of lineality
• The expectation of a linear combination of randomvariables is the linear combination of expectations.
( ) ( )( )
( ) ( )
)()(
)()(
)()(
)()()(
Y E X E
P Y P X
P Y X Y X E
Y E X E Y X E
βα
ωωβωωα
ωωβωαβα
βαβα
ωω
ω
+=
=+=
+=+
+=+
∑∑
∑
Ω∈Ω∈
Ω∈
21
Finding the maximum
• Suppose that n children of differing heights areplaced in line at random. You select the first
child from the line and walk with her/him alongthe line until you encounter a child who is talleror until you have reached the end of the line. If
you do encounter a taller child, you repeat theprocess.
• What is the expected value of the number of
children selected from the line?
Taken fromTijms, Understanding probability22
Finding the maximum
• We define the variable X as the number ofchildren selected from the line.
• We will define the indicator variable:
• Now the number of selected children will
be:
= otherwise 0
linethefromselectedischildtheif 1 i-th X i
n X X X X +++= L21
23
Finding the maximum
• The probability that the ith child is thetallest among the first i children is 1/i
• Therefore:
( ) n1,2,ifor11
11
10 L==⋅+
−⋅=
iii X E i
( ) ( )
( ) 57722.02
1)ln(
1
3
1
2
11
21
++≈+++=
+++=
nn
n X E
X X X E X E n
L
L
24
Expected number of distinct
birthdays• What is the expected number of distinct
birthdays within a randomly formed groupof 100 persons.
– We define the random variable
– The number of birthdays is
= otherwise 0
dayonis birthdaytheif 1 i X i
36 521 X X X X +++= L
Taken fromTijms, Understanding probability
8/17/2019 Small Mathematical Expectation
5/6
25
Expected number of distinct
birthdays
• What is the expected number of distinct birthdays withina randomly formed group of 100 persons.
– For each day we have:
– The expected number of distinct birthdays is
)0(1)1(
365
364)0(
100
=−==
==
ii
i
X P X P
X P
Taken fromTijms, Understanding probability
( ) ( ) 6.87365
3641365
10 0
36 521 =
−=+++= X X X E X E L
26
Expected number of distinct
birthdays
• For an arbitrary number of persons.
( )
−=
n
X E 365
3641365
50 100 150 200 250 300 350
50
100
150
200
250
300
350
Number of persons
E x p e c t e d n u m b e r o f d i s t i n c t b i r t h d a y s
25 30 35 40 45
20
25
30
35
40
45
50
27
Other Properties
• If X is a non negative random variable,then
• Also
0)( ≥ X E
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
1 6 1 1
1 6
2 1
2 6
3 1
3 6
4 1
4 6
5 1
5 6
6 1
)()( then Y E X E Y X ≥≥
28
Other Properties
• What do we mean by
?)()( then Y E X E Y X ≥≥
ℜ→Ω: X
0 1 X(.) X(.)
Y(.) Y(.)
X
0 1 P(X) P(X)
P(Y) P(Y)
X
ℜ→Ω:Y
X
∑∈
== X x
x X xP X E )()(
29
Caveats of intuition
• Does the expectation always exist?
∑∈
= X x
x N
X Mean1
)(
∑∈
== X x
x X xP X E )()(
http://physicsweb.org/articles/world/14/7/9/1#pw1407091
The physics of the WebJuly 2001
λλ −== e x
x X P x
!)(
α x
x X P 1
)( ==
Sample mean always exists
Expectation
perhaps givesan infinite
value !!!!
30
Caveats of intuition
• Does the expectation always exist?
• Example: A Cauchy random variable
∑∈
= X x
x N
X Mean1
)(
∑∈
== X x
x X xP X E )()(
http://physicsweb.org/articles/world/14/7/9/1#pw1407091
The physics of the WebJuly 2001
( ) ( )+∞
8/17/2019 Small Mathematical Expectation
6/6
31
Caveats of intuition
• Why is infinite?
• Divergent series
• Note that for high values of x
( )∑∑ ∞
=∈ +===
1
21
)()( x X x x
x x X xP X E
π
α x x X P
1)( ==
( ) ( )
∑
∑ ∑∑∞
=
∞
=
∞
>>
>>
=
∝
++
≅+
=
1
1 1
1
1
22
1)(
1111
)(2
2
x
x x
x
x
x X E
x x x
x x X E
πππ
32
Caveats of intuition
• Why is infinite?
• Value of the harmonic series
0 10 00 200 0 3 00 0 4 000 5 000 6 000 7 00 0 8 00 0 900 0 100005
5.5
6
6.5
7
7.5
8
8.5
9
9.5
10
∑=
N
x x1
1
∞=+++>
++++++++>
++++++++=
∑
∑
∑
∞
=
∞
=
∞
=
L
L
L
1 1 1 11
8
1
8
1
4
1
4
1
4
1
4
1
2
1
2
11
1
91
81
71
61
51
41
31
2111
1
1
1
x
x
x
x
x
x REPASSAR
33
Expected waiting times
• Geometric
• Pareto
• gausian
34
Expected value of a Binomial