[SciPy-user] implementation of multi-objective genetic algorithm in scipy

josef.pktd@gmai... josef.pktd@gmai...
Fri Mar 6 15:24:42 CST 2009


On Fri, Mar 6, 2009 at 3:59 PM, Scott Bucking <sbucking@gmail.com> wrote:
> Hi,
>
> Anyone familiar with a multi-objective genetic algorithm (MOGA)
> written in python?  I saw a note on the scipy website that a GA
> existed (under the optimize package).  Does this kit include a MOGA?
>
> I can't find any reference to a GA in the documentation or source code.
>
> Cheers,

scipy.optimize has simulated annealing  `anneal`
there is a genetic algorithm that has been moved to the `learn`
scikits, but I never used it, and it hasn't been modified in 3 years.

http://scipy.org/scipy/scikits/browser/trunk/learn/scikits/learn/machine/ga

I don't think there is a MOGA. There are several other machine
learning packages for python, but I don't know if any of them contains
more.

If your optimization is computationally not very expensive, then you
could trace the pareto frontier (of the convex hull of the feasible
payoff set) by varying linear weight parameters of the objective
functions.

Josef


More information about the SciPy-user mailing list