Vista wrote:
> It looks awfully like some type of optimization problem and/or some
types of
> datamining problem, such as sup****t vector machines, etc. I am sure
there
> are simple and elegant algorithms handling such problem. I just don't
know
> where to find them.
Expanding my longer answer posted elsewhere, yes, this could be solved
by an SVM. That changes the LP to a quadratic program but pretty much
everything else I wrote still applies. An advantage of an SVM over the
LP approach is that the SVM is probably more efficient at fitting higher
order (polynomial) models and certain other nonlinear models. (A
polynomial model can be viewed as a linear model in a higher dimensional
space, in which the original space is embedded.) If you're sure the
separator is linear in the original space, though, I'm not sure an SVM
gives any huge advantage over an LP (or at least the "right" LP model,
whatever that might turn out to be).
/Paul
|
