> |
MATH 392 -- Seminar in Computational Commutative Algebra
Maximum Likelihood Estimation in a Linear Mixture Model
We wish to determine the values of that maximize
the log-likelihood function
l =
where
and from the data
> | ![]() |
> | ![]() |
> | ![]() |
> | ![]() |
The log-likelihood function is
> | ![]() |
To maximize, we apply the usual process from multivariable calculus:
> | ![]() |
> | ![]() |
To solve the equationswe will use our Grobner basis tools:
> | ![]() |
> | ![]() |
> | ![]() |
> | ![]() |
> | ![]() |
Of these, note that only one is in the range
> | ![]() |
We have exactly one root in the "probability simplex"
Since is relatively large here, the data would indicate
that this region is a CG-rich area.
> |