Download Week4-1

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts
no text concepts found
Transcript
Moment generating function
The moment generating function of random variable X is
given by
 (t )  E[etX ]
Moment generating function
The moment generating function of random variable X is
given by
 (t )  E[etX ]
 (t )   etx p( x) if X is discrete
x
Moment generating function
The moment generating function of random variable X is
given by
 (t )  E[etX ]
 (t )   etx p( x) if X is discrete
x

 (t )   etx f ( x)dx if X is continuous

dE[etX ]
d tX
 '(t ) 
 E[ e ]  E[ XetX ]
dt
dt
  '(0)  E[ X ]
dE[etX ]
d tX
 '(t ) 
 E[ e ]  E[ XetX ]
dt
dt
  '(0)  E[ X ]
d
 (t )  E[ XetX ]  E[ X 2 etX ]
dt
  (2) (0)  E[ X 2 ]
(2)
More generally,
d k 1 tX
 (t )  E[ X e ]  E[ X k etX ]
dt
  ( k ) (0)  E[ X k ]
(k )
Example: X has the Poisson distribution with parameter l
Example: X has the Poisson distribution with parameter l
x l
x l
(
l
)
e
(
l
)
e

tX
tx
tx
 (t )  E[e ]   x 0 e
  x 0 e
x!
x!
t x
t
(
l
e
)

l
 l l et
l ( ee 1)
=e  x 0
e e e
x!

Example: X has the Poisson distribution with parameter l
x l
x l
(
l
)
e
(
l
)
e

tX
tx
tx
 (t )  E[e ]   x 0 e
  x 0 e
x!
x!
t x
(
l
e
)

l
 l l et
l ( et 1)
=e  x 0
e e e
x!

  (t ) 
d [e
l ( et 1)
dt
]
 le e
t l ( et 1)
Example: X has the Poisson distribution with parameter l
x l
x l
(
l
)
e
(
l
)
e

tX
tx
tx
 (t )  E[e ]   x 0 e
  x 0 e
x!
x!
t x
(
l
e
)

l
 l l et
l ( et 1)
=e  x 0
e e e
x!

  (t ) 
d [e
l ( et 1)
dt
]
 l et el ( e 1)
t
ax
a

e
 x 0 x!

t l ( et 1)
 '(0)  l e e
t 0
l
t l ( et 1)
 '(0)  l e e
t l ( et 1)
 ''(0)  l e e
t 0
l
t l ( et 1)
 le le e
t
t 0
 l  l2
 '(0)  l e e
t l ( et 1)
t l ( et 1)
 ''(0)  l e e
t 0
l
t l ( et 1)
 le le e
t
 Var ( X )  E[ X 2 ]  E[ X ]2  l
t 0
 l  l2
If X and Y are independent, then
 X Y (t )  E[et ( X Y ) ]  E[etX etY ) ]  E[etX ]E[etY ]
= X (t )Y (t )
The moment generating function of the sum of two random
variables is the product of the individual moment generating
functions
Let Y = X1+X2 where X1~Poisson(l1) and X2~Poisson(l2) and
X1 and X1 are independent, then
E[etY ]  E[et ( X1  X 2 ) ]  E[etX1 ]E[etX 2 ]
=e
l1 ( et 1) l2 ( et 1)
e
e
( et 1)( l1  l2 )
Let Y = X1+X2 where X1~Poisson(l1) and X2~Poisson(l2) and
X1 and X1 are independent, then
E[etY ]  E[et ( X1  X 2 ) ]  E[etX1 ]E[etX 2 ]
=e
l1 ( et 1) l2 ( et 1)
e
 Y ~ Poisson(l1  l2 )
e
( et 1)( l1  l2 )
Note: The moment generating function uniquely determines
the distribution.
Markov’s inequality
If X is a random variable that takes only nonnegative values,
then for any a > 0,
E[ X ]
P( X  a) 
.
a
Proof (in the case where X is continuous):

E[ X ]   xf ( x)dx

a


a
  xf ( x)dx   xf ( x)dx

  xf ( x)dx
a


a
a
  af ( x)dx  a  f ( x)dx  aP( X  a)
Strong law of large numbers
Let X1, X2, ..., Xn be a set of independent random variables
having a common distribution, and let E[Xi] = m. then, with
probability 1
X1  X1  ...  X n
 m as n  .
n
Central Limit Theorem
Let X1, X2, ..., Xn be a set of independent random variables
having a common distribution with mean m and variance s.
Then the distribution of
X 1  X 1  ...  X n  nm
s n
tends to the standard normal as n  . That is
P(
X 1  X 1  ...  X n  nm
as n  .
s n
 a) 
1
2

a

e
 x2 / 2
dx
Conditional probability and
conditional expectations
Let X and Y be two discrete random variables, then the
conditional probability mass function of X given that Y=y is
defined as
P{ X  x, Y  y} p( x, y)
pX |Y ( x | y)  P{ X  x | Y  y} 

.
P{Y  y}
p( y )
for all values of y for which P(Y=y)>0.
Conditional probability and
conditional expectations
Let X and Y be two discrete random variables, then the
conditional probability mass function of X given that Y=y is
defined as
P{ X  x, Y  y} p( x, y)
pX |Y ( x | y)  P{ X  x | Y  y} 

.
P{Y  y}
p( y )
for all values of y for which P(Y=y)>0.
The conditional expectation of X given that Y=y is defined as
E[ X | Y  y ]   xP{ X  x | Y  y}   xp X |Y ( x | y ).
x
x
Let X and Y be two continuous random variables, then the
conditional probability density function of X given that Y=y
is defined as
f ( x, y)
f X |Y ( x | y) 
.
fY ( y )
for all values of y for which fY(y)>0.
The conditional expectation of X given that Y=y is defined as

E[ X | Y  y]   xf X |Y ( x | y)dx.

E[ X ]  E[ E[ X | Y  y ]]  E[ E[ X | Y ]]
E[ X ]   E[ X | Y  y ]P(Y  y ) if Y is discrete
y

E[ X ]   E[ X | Y  y ] f ( y )dy if Y is continuous.

Proof:
Related documents