Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Scribd
Upload
8 views

Lecture 2

mat stat

Uploaded by

katrineenghoff
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
8 views

Lecture 2

mat stat

Uploaded by

katrineenghoff
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 17

ST521: Mathematical Statistics

Random variables Part I

Jing Qin

ST521 (IMADA) Mathematical Statistics: Lecture 2 1 / 12


Outline

1 General definition

2 Discrete random variables

ST521 (IMADA) Mathematical Statistics: Lecture 2 2 / 12


(+) Random variable and distribution function

(Def +) A random variable is a function X : Ω → R with the


property that {ω|X (ω) ≤ x} ∈ F for each x ∈ R. Such function is
said to be F-measurable.
(Def 4.1+) The distribution function of a random variable X is the
function F : R → [0, 1] given by

FX (x) = P(X ≤ x).

Example: (Exp 2.14)


Flip a (balanced) coin twice, we could consider the following:
1 Ω = {HH, HT , TH, TT }, F =? and P(A) =?
2 X := the number of heads observed.
3 X ({HH}) = 2, X ({HT }) = X ({TH}) = 1 and X ({TT }) = 0.
4 Question: what is the distribution function of the random variable
X ? Draw a diagram with x values as x-axis and F (x) in y-axis.

ST521 (IMADA) Mathematical Statistics: Lecture 2 3 / 12


Categories of distributions
(Def 2.12+) A random variable X is discrete if its range contains
only at most countable values in R. (Exp 2.14)
(Def 4.2) A random variable X is continuous if its distribution
function is continuous. E.g. throw an arrow on the floor. Consider
its angle when landing.
(+) There can be random variables that are a mixture between
discrete and continuous random variables. E.g. Toss a coin+throw
arrow depending on the coin result.

ST521 (IMADA) Mathematical Statistics: Lecture 2 4 / 12


Outline

1 General definition

2 Discrete random variables

ST521 (IMADA) Mathematical Statistics: Lecture 2 5 / 12


Probability (mass) function p(y )

(Def 3.2) Probability (mass) function of a r.v. Y is the following


p(y ) := P(Y = y ) =sum of the probabilities of sample points that
are assigned value y .
P
(Thm 3.1) 0 ≤ p(y ) ≤ 1 and y p(y ) = 1.
(Revisit) Example: (Exp 2.14)
Flip a (balanced) coin twice (first time, second time), we could
consider the following:
1 Ω = {HH, HT , TH, TT };
2 X := the number of heads observed.
3 X ({HH}) = 2, X ({HT }) = X ({TH}) = 1 and X ({TT }) = 0.
4 Exercise: Calculate p(2), p(1) and p(0).

ST521 (IMADA) Mathematical Statistics: Lecture 2 6 / 12


Expectation of r.v.
Given pY (y ) = P(Y = y ).
P P
(Def 3.4) If y |y | · pY (y ) < ∞, then E(Y ) = y y · pY (y ) is called
the expectation/expected value/mean of the r.v. Y . It tells the
“central value" of Y .
(Thm 3.2) Let g(Y ) be a real-valued function of Y and pY (y ) be
the probability function of Y . Then the expected value E(g(Y ))
has the following property
X
E[g(Y )] = g(yj ) · pY (yj ).
yj

This tells us to calculate E(g(Y )), we only need to know pY


instead of pg(Y ) .
Proof.

ST521 (IMADA) Mathematical Statistics: Lecture 2 7 / 12


Variance of r.v.

(Def 3.5) If Y is a r.v. with E(Y ) = µ. The variance of Y is defined


to be
V (Y ) = E[(Y − µ)2 ].

Informally, it measures how far (Euclidean distance) the values of


Y spread away from µ.
Sometimes using notation σ 2 .
The standard deviation of Y is the positive square root of V (Y ).
Example: Calculate µ and σ 2 for the following r.v. Y :
y 0 1 2 3
p(y ) 1/8 1/4 3/8 1/4
µ = 1.75 and σ 2 = .9375

ST521 (IMADA) Mathematical Statistics: Lecture 2 8 / 12


Variance of r.v.

(Def 3.5) If Y is a r.v. with E(Y ) = µ. The variance of Y is defined


to be
V (Y ) = E[(Y − µ)2 ].

Informally, it measures how far (Euclidean distance) the values of


Y spread away from µ.
Sometimes using notation σ 2 .
The standard deviation of Y is the positive square root of V (Y ).
Example: Calculate µ and σ 2 for the following r.v. Y :
y 0 1 2 3
p(y ) 1/8 1/4 3/8 1/4
µ = 1.75 and σ 2 = .9375

ST521 (IMADA) Mathematical Statistics: Lecture 2 8 / 12


Properties of E(Y )

Let Y be a random variable in the following.


(Thm 3.3) Let c be a constant, then E(c) = c.
Proof.

ST521 (IMADA) Mathematical Statistics: Lecture 2 9 / 12


Properties of E(Y )

Let Y be a random variable in the following.


(Thm 3.3) Let c be a constant, then E(c) = c.
Proof.

ST521 (IMADA) Mathematical Statistics: Lecture 2 9 / 12


Properties of E(Y )

(Thm 3.4) Let g(Y ) be a function of Y and c is a constant, then


E[c · g(Y )] = c · E[g(Y )].
Proof.

ST521 (IMADA) Mathematical Statistics: Lecture 2 10 / 12


Properties of E(Y )

(Thm 3.4) Let g(Y ) be a function of Y and c is a constant, then


E[c · g(Y )] = c · E[g(Y )].
Proof.

ST521 (IMADA) Mathematical Statistics: Lecture 2 10 / 12


Linearity of E(Y )

(Thm 3.5) Let g1 (Y ), g2 (Y ) be functions of Y , then


E[g1 (Y ) + g2 (Y )] = E[g1 (Y )] + E[g2 (Y )].
Proof.
Thm 3.4+3.5 can give rise to the linearity of E(Y ). Let
g1 (Y ), g2 (Y ), . . . , gk (Y ) be functions of Y and c1 , c2 , . . . , ck be
some constants. Then we have
" k # k
X X
E ci · gi (Y ) = ci · E[gi (Y )].
i=1 i=1

Later
Pnwe will seePanother result Thm (5.12),
E( i=1 ai Yi ) = ni=1 ai · E(Yi ) within multi-variate probability
distribution.

ST521 (IMADA) Mathematical Statistics: Lecture 2 11 / 12


Linearity of E(Y )

(Thm 3.5) Let g1 (Y ), g2 (Y ) be functions of Y , then


E[g1 (Y ) + g2 (Y )] = E[g1 (Y )] + E[g2 (Y )].
Proof.
Thm 3.4+3.5 can give rise to the linearity of E(Y ). Let
g1 (Y ), g2 (Y ), . . . , gk (Y ) be functions of Y and c1 , c2 , . . . , ck be
some constants. Then we have
" k # k
X X
E ci · gi (Y ) = ci · E[gi (Y )].
i=1 i=1

Later
Pnwe will seePanother result Thm (5.12),
E( i=1 ai Yi ) = ni=1 ai · E(Yi ) within multi-variate probability
distribution.

ST521 (IMADA) Mathematical Statistics: Lecture 2 11 / 12


Application: Property of V (Y )

(Thm 3.6) Let E(Y ) = µ, then


V [Y ] = E[Y 2 ] − (E(Y ))2 = E[Y 2 ] − µ2 .
We also have V (Y ) = E[Y (Y − 1)] + E(Y ) − [E(Y )]2 (Useful in
the next lecture).
Proof.

ST521 (IMADA) Mathematical Statistics: Lecture 2 12 / 12


Application: Property of V (Y )

(Thm 3.6) Let E(Y ) = µ, then


V [Y ] = E[Y 2 ] − (E(Y ))2 = E[Y 2 ] − µ2 .
We also have V (Y ) = E[Y (Y − 1)] + E(Y ) − [E(Y )]2 (Useful in
the next lecture).
Proof.

ST521 (IMADA) Mathematical Statistics: Lecture 2 12 / 12

You might also like