1. 1
METHOD OF MOMENTS :
If E(Xi) is some function of , then we can estimate that
function () using only the statistic x
We know how to estimate parameters such as , 2, or p. Now we
need general methods to estimate other parameters
From a sample X1,…,Xn, suppose we wish to estimate some
unknown parameter
Estimating Parameters
2. 2
Example: X1,…,Xn ~Exponential()
E(X)=1/ (so ()=1/ )
X
1
ˆ
)
ˆ
(
X
Now try to solve for as a function of , and call the
solution ̂
x
Method of Moments Estimator
Method of Moments : Solve for :
3. 3
Suppose there are two parameters to estimate (, ) and that
)
S
,
X
(
of
functions
are
)
ˆ
,
ˆ
(
o
S
)
ˆ
,
ˆ
h(
S
)
ˆ
,
ˆ
g(
X
that
so
ˆ
,
ˆ
find
s
MME'
2
2
E(X)= g(, ) Var(X)= h(, )
Key: solve 2 eqns
for 2 unknowns
Method of Moments Estimator
8. 8
If the probability density function for X is f(x; ), then if
sample items are obtained independently, the p.d.f. for
(X1,…,Xn) is (the joint density function)
f(x1,…xn; )= f(x1; )f(x2; )f(xn; )
Maximum Likelihood Estimation
We usually look at f as a function of (X1,…,Xn), but it is
also a function of the parameter . That is, we can
write L(; x1,…xn)= f(x1; )f(x2; )f(xn; )
9. 9
Maximum Likelihood Estimation
)
|
f(x
)
|
x
f( i
n
1
i
~
The MAXIMUM LIKELIHOOD ESTIMATOR (MLE)
of is the value of that maximizes L( | x )
L(; x1,…xn) is called the LIKELIHOOD of the data if we
write it as a function of parameters
)
|
f(x
)
x
|
L( i
1
~
n
i
10. 10
Hint : it is almost always easier to maximize the log of L rather
than L, and it maximizes at the same value of
x
-
n
1 e
f(x)
so
)
Exp(
~
X
,...,
X
Note : Check second derivative to make sure you didn’t find the
minimum likelihood estimator
MLE of for Exponential
)
x
exp(-
e
)
x
|
L(
1
i
n
x
-
1
~
i
n
i
n
i
12. 12
1) Identify f(x|) :
Computing the MLE: 6 Steps
)
x
(
f i
n
1
i
L() =
3) Write this as a function of the parameter L() :
2) Using data x1,…xn from the distribution in 1)
write the joint distribution f(x1,x2,…,xn| )
13. 13
4) To find the value of that maximizes
L()= L(|x1,…xn),examine the log (ln L()) instead
Computing the MLE: 6 Steps
̂
6) Solve this equation for , call it . It is the MLE if the
second derivative is negative
0)
)
L(
ln
(
5) Find max by taking derivative, set it to zero
17. 17
*
X
ˆ
Same thing as I(0 < x* < ), where x* = Max(x1, x2 ,…, xn).
To make L() as Large as possible, choose to be as small as
possible, but not so small that < x* (otherwise L=0).
MLE for Uniform Distribution