Scribd
Upload
42 views2 pages

Comparison To ML Estimation

The document compares maximum likelihood (ML) and maximum a posteriori (MAP) estimation methods. ML estimation finds the value of x that maximizes the likelihood function fY|X(y|x), while MAP estimation maximizes the joint distribution fY|X(y|x)fX(x). For some distributions like a uniformly distributed prior, the ML and MAP estimates will be the same. As an example, when a signal X is transmitted over a channel with noise W, the ML estimate of X given an observation Y=y is simply y, while the MAP estimate is a weighted average of y and the prior mean of X.

Uploaded by

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

Comparison To ML Estimation

The document compares maximum likelihood (ML) and maximum a posteriori (MAP) estimation methods. ML estimation finds the value of x that maximizes the likelihood function fY|X(y|x), while MAP estimation maximizes the joint distribution fY|X(y|x)fX(x). For some distributions like a uniformly distributed prior, the ML and MAP estimates will be the same. As an example, when a signal X is transmitted over a channel with noise W, the ML estimate of X given an observation Y=y is simply y, while the MAP estimate is a weighted average of y and the prior mean of X.

Uploaded by

bossishere
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

Comparison to ML Estimation https://www.probabilitycourse.com/chapter9/9_1_3_comparison_to_M...

9.1.3 Comparison to ML Estimation


We discussed maximum likelihood estimation in the previous chapter. Assuming that we have observed
Y = y, the maximum likelihood (ML) estimate of X is the value of x that maximizes
fY|X (y|x) (9.1)
̂  . On the other hand, the MAP estimate of X is the value of x that
We show the ML estimate of X by xML
maximizes

fY|X (y|x)fX (x) (9.2)


The two expressions in Equations 9.1 and 9.2 are somewhat similar. The difference is that Equation 9.2 has
an extra term, fX (x). For example, if X is uniformly distributed over a finite interval, then the ML and the
MAP estimate will be the same.

Example 9.5
Suppose that the signal X ∼ N(0, σX2 ) is transmitted over a communication channel. Assume that the
received signal is given by

Y = X + W,
where W ∼ N(0, σW2 ) is independent of X .
1. Find the ML estimate of X , given Y = y is observed.
2. Find the MAP estimate of X , given Y = y is observed.

• Solution
◦ Here, we have
2
1 − x2
fX (x) = e 2σX .
√2π
‾ σX
‾‾
We also have, Y|X =x ∼ N(x, σW2 ), so
(y−x)2
1 − 2
fY|X (y|x) = e 2σW
.
√2π
‾ σW
‾‾
1. The ML estimate of X , given Y = y, is the value of x that maximizes
(y−x)2
1 − 2
fY|X (y|x) = e 2σW
.
√2π
‾ σW
‾‾

1 of 2 10/12/22, 3:53 pm
Comparison to ML Estimation https://www.probabilitycourse.com/chapter9/9_1_3_comparison_to_M...

To maximize the above function, we should minimize (y − x)2 . Therefore, we conclude


̂  = y.
xML
2. The MAP estimate of X , given Y = y, is the value of x that maximizes

{ [ 2σW2 2σX2 ]}
(y − x)2 x2
fY|X (y|x)fX (x) = c exp − + ,

where c is a constant. To maximize the above function, we should minimize

(y − x)2 x2
+ .
2σW2 2σX2
By differentiation, we obtain the MAP estimate of x as

σX2
̂ 
xMAP = y.
σX2 + σW2

← previous
next →

The print version of the book is available through Amazon here.

2 of 2 10/12/22, 3:53 pm

You might also like