You could say that the LLN and CLT are conceptually same. But that
statement is not quite correct as the LLN deals with observations of
a certain random variable approaching the expected value of the
random variable as the number of sampled observations increase where
as the CLT is entirely for RVs explaining that as you continue taking
more and more independent and identically distributed (IID) random
variables with specified mean and variance values, the averaged sum
will eventually be normally distributed.

Min/max estimations on the last rely just means minimization or
maximization estimation problem.

......Khalid Gidaya.

. Please look for more info on wikipediaboth

--- In i..., "Genial Fido"
wrote:
>
> Hi All,
>
> Thanks for your suggestions.
>
> I want to know the difference between law of large number and
Central limit
> theorem. Conceptually both are same, but the LLN is applied on the
number
> and CLT is applied on Gaussian distribution or in general on any
> distributions. Is this correct ?
>
> please can some body explain me Min/max estimations
>
> thanking in advance
> fido
>
> On Mon, Jul 28, 2008 at 4:04 PM, Mamo wrote:
>
> > Estimation problems are everywhere in Signal/Image processing
> > applications and Gaussian models are the best PSF in representing
> > many natural processes relating to Signals and systems. Consider
> > image degradation problem caused by atmospheric bluring of an
image
> > by a Gausian like PSF--Closely related to the heat equation. When
> > trying to estimate an unobservable process from its model Random
> > variable, the Gausian fits many real life problems. Isn't difusion
> > very common phenomenon? Well, ones you model your data in a
Gausian,
> > you will have simpler, closed form recursive solution to your
Min/Max
> > estimations problem using methods like MLEM, PML, or recursive
Kalman
> > filter.......Khalid Gidaya.
> >
> >
> > --- In i... ,
ramachan@
> > wrote:
> > >
> > > gaussian models are useful because they lead to mathematically
> > simple
> > > solutions, with most often closed form solutions.
> > >
> > >
> > > Quoting fido.genial@:
> > >
> > > > Hi All,
> > > >
> > > > I might be asking trivial question,but i am forced to ask. I
> > need
> > > > to know why is it most of the application in signals
processing
> > and
> > > > image processing use Gaussian models . For example in image
> > pattern
> > > > recoginzation and in Background subtraction which uses
Gaussian
> > > > Mixture model. Please can someone help me out with this.
> > > >
> > > > Thanks in advance
> > > > Fido
> > > >
> > >
> >
> >
>

Hi All,

Thanks for your suggestions.

I want to know the difference between law of large number and Central limit
theorem. Conceptually both are same, but the LLN is applied on the number
and CLT is applied on Gaussian distribution or in general on any
distributions. Is this correct ?

please can some body explain me Min/max estimations

thanking in advance
fido

On Mon, Jul 28, 2008 at 4:04 PM, Mamo wrote:

> Estimation problems are everywhere in Signal/Image processing
> applications and Gaussian models are the best PSF in representing
> many natural processes relating to Signals and systems. Consider
> image degradation problem caused by atmospheric bluring of an image
> by a Gausian like PSF--Closely related to the heat equation. When
> trying to estimate an unobservable process from its model Random
> variable, the Gausian fits many real life problems. Isn't difusion
> very common phenomenon? Well, ones you model your data in a Gausian,
> you will have simpler, closed form recursive solution to your Min/Max
> estimations problem using methods like MLEM, PML, or recursive Kalman
> filter.......Khalid Gidaya.
> --- In i... , ramachan@...
> wrote:
> >
> > gaussian models are useful because they lead to mathematically
> simple
> > solutions, with most often closed form solutions.
> >
> >
> > Quoting fido.genial@...:
> >
> > > Hi All,
> > >
> > > I might be asking trivial question,but i am forced to ask. I
> need
> > > to know why is it most of the application in signals processing
> and
> > > image processing use Gaussian models . For example in image
> pattern
> > > recoginzation and in Background subtraction which uses Gaussian
> > > Mixture model. Please can someone help me out with this.
> > >
> > > Thanks in advance
> > > Fido
> > >
> >
>

Estimation problems are everywhere in Signal/Image processing
applications and Gaussian models are the best PSF in representing
many natural processes relating to Signals and systems. Consider
image degradation problem caused by atmospheric bluring of an image
by a Gausian like PSF--Closely related to the heat equation. When
trying to estimate an unobservable process from its model Random
variable, the Gausian fits many real life problems. Isn't difusion
very common phenomenon? Well, ones you model your data in a Gausian,
you will have simpler, closed form recursive solution to your Min/Max
estimations problem using methods like MLEM, PML, or recursive Kalman
filter.......Khalid Gidaya.
--- In i..., ramachan@... wrote:
>
> gaussian models are useful because they lead to mathematically
simple
> solutions, with most often closed form solutions.
> Quoting fido.genial@...:
>
> > Hi All,
> >
> > I might be asking trivial question,but i am forced to ask. I
need
> > to know why is it most of the application in signals processing
and
> > image processing use Gaussian models . For example in image
pattern
> > recoginzation and in Background subtraction which uses Gaussian
> > Mixture model. Please can someone help me out with this.
> >
> > Thanks in advance
> > Fido
>

gaussian models are useful because they lead to mathematically simple
solutions, with most often closed form solutions.
Quoting f...@gmail.com:

> Hi All,
>
> I might be asking trivial question,but i am forced to ask. I need
> to know why is it most of the application in signals processing and
> image processing use Gaussian models . For example in image pattern
> recoginzation and in Background subtraction which uses Gaussian
> Mixture model. Please can someone help me out with this.
>
> Thanks in advance
> Fido
>

i guess Central Limit Theorem..

Hi All,
>
> I might be asking trivial question,but i am forced to ask. I need to know why is it most of the application in signals processing and image processing use Gaussian models . For example in image pattern recoginzation and in Background subtraction which uses Gaussian Mixture model. Please can someone help me out with this.
>
>Thanks in advance
>Fido
>
>

Hi All,

I might be asking trivial question,but i am forced to ask. I need to know why is it most of the application in signals processing and image processing use Gaussian models . For example in image pattern recoginzation and in Background subtraction which uses Gaussian Mixture model. Please can someone help me out with this.

Thanks in advance
Fido