FPGA Central

I also have question on basic concept difference between likelihood
and probability. Thanks
I have confused this for a while

dgse said the following on 10/05/2007 06:31:
> I also have question on basic concept difference between likelihood
> and probability. Thanks
> I have confused this for a while

If x is the parameter we're trying to estimate, and y is our
observation, then:

ML is:

argmax f(y|x)
x

MAP is:

argmax f(x|y)
x

ML assumes nothing about the prior distribution of x (it will even work
if x is deterministic). MAP requires prior knowledge of the
distribution of x.


--
Oli

On May 10, 3:11 am, Oli Charlesworth <c...@olifilth.co.uk> wrote:
> dgse said the following on 10/05/2007 06:31:
>
> > I also have question on basic concept difference between likelihood
> > and probability. Thanks
> > I have confused this for a while
>
> If x is the parameter we're trying to estimate, and y is our
> observation, then:
>
> ML is:
>
> argmax f(y|x)
> x
>
> MAP is:
>
> argmax f(x|y)
> x
>
> ML assumes nothing about the prior distribution of x (it will even work
> if x is deterministic). MAP requires prior knowledge of the
> distribution of x.
>
> --
> Oli

And then using Bayes' rule, the original poster can derive ..... ;-)