Scribd
Upload
3 views2 pages

Maximum Likelihood Estimation for Normals

The document outlines the maximum likelihood estimators (MLEs) for parameters a, b, c, and σ² derived from four normal populations with the same variance. It provides the likelihood function and the simultaneous equations necessary for estimating these parameters based on sample observations. The final MLEs are expressed in terms of the sample means and the sample size n.

Uploaded by

tdharmender369
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
3 views2 pages

Maximum Likelihood Estimation for Normals

The document outlines the maximum likelihood estimators (MLEs) for parameters a, b, c, and σ² derived from four normal populations with the same variance. It provides the likelihood function and the simultaneous equations necessary for estimating these parameters based on sample observations. The final MLEs are expressed in terms of the sample means and the sample size n.

Uploaded by

tdharmender369
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd

Practical-7

Statement: A sample of size n is drawn from each of the four normal populations which have the
same variance σ 2 . The means of the four populations are a+ b+c , a+b−c , a−b+c and a−b−c .
What are the [Link]. for a, b, c and σ 2 ?

Formula Used:

The MLE’s for a, b, c and σ 2 are the solutions of the simultaneous equations (maximum likelihood
equations for estimating a, b, c and σ 2 ¿:

∂ ∂
log L=0 … (1) log L=0 … (2)
∂a ∂b
∂ ∂
log L=0 … (3) 2
log L=0 … (4)
∂c ∂σ

Calculations:

Let the sample observations be denoted by x ij, i=1 , 2 ,3 , 4 ; j=1 , 2 ,… ,n . Since the four samples,
from the four normal populations are independent, the likelihood function L of all the observations
x ij, ( i=1 , 2 ,3 , 4 ; j=1 , 2 ,… ,n ) , is given by:

( ) { }
4π 4 n
1 −1
2 ∑ ∑ ( ij
L= exp x −μi )2 ,
σ √2 π 2σ i=1 j=1

Where μi ,(i=1 , 2 ,3 , 4 ) is mean of the ith population. Therefore,

( )
4π
1
L= exp ¿ ¿
σ √2 π

1
2 ∑ ( 1j
x −a−b−c ) +¿ ∑ ( x 2 j−a−b+c ) +¿ ∑ ( x 3 j−a+b−c ) +¿ ∑ ( x 4 j−a+ b+c )
2 2 2 2 2
log L=k−2 nlog σ − {
2σ j j j j

Where k is a constant w.r. to a, b, c and σ 2. The MLE’s for a, b, c and σ 2 are the solutions of the
simultaneous equations (maximum likelihood equations for estimating a, b, c and σ 2 ¿:

∂ ∂
log L=0 … (1) log L=0 … (2)
∂a ∂b
∂ ∂
log L=0 … (3) 2
log L=0 … (4)
∂c ∂σ
(1) gives:

−1
2
¿
2σ
∑ ( x1 j+ x 2 j+ x 3 j+ x 4 j ) +n [ (−a−b−c ) +(−a−b+ c )+ (−a+b−c ) +(−a+b+ c ) ]=0
j

∑ ¿¿
j=1

4 n
1
a^ = ∑ ∑ x =x
4 n i=1 j=1 ij

2) gives:

−1
2
¿
2σ

∑ ( x1 j+ x 2 j−x 3 j−x 4 j ) +n [ (−a−b−c ) +(−a−b+ c ) −(−a+ b−c ) −(−a+b +c ) ]=0


j

∑ x 1 j +∑ x 2 j −∑ x 3 j−∑ x 4 j−4 nb=0


j j j j

1
^
∴ b=
4n (∑ x j
1j +∑ x 2 j−∑ x 3 j −∑ x 4 j
j j j
)
x 1 + x 2−x 3−x 4
^
b= , where x i is the mean of the Ith sample.
4
x 1−x 2+ x3 −x 4
Similarly (3) will give: c^ =
4
(4) gives:

−2 n 1
2
+ 4¿
σ 2σ
2 1
∴ σ^ = ¿
4n
Conclusion:

The [Link] for a, b, c and σ 2 are:


4 n
1 x 1 + x 2−x 3−x 4 x 1−x 2+ x3 −x 4
a^ = ∑ ∑ x ij =x , b=
^ c^ =
4 n i=1 j=1 4 4

2 1
σ^ = ¿
4n

You might also like