Print

Print


No.  Tutorial4 uses the standard koza.params file, which doesn't even  
have mutation attached in the pipeline.  Specifically:

pop.subpop.0.species.pipe = ec.breed.MultiBreedingPipeline
# Koza's decision here was odd...
pop.subpop.0.species.pipe.generate-max = false
# Subsidiary pipelines:
pop.subpop.0.species.pipe.num-sources = 2
pop.subpop.0.species.pipe.source.0 = ec.gp.koza.CrossoverPipeline
pop.subpop.0.species.pipe.source.0.prob = 0.9
pop.subpop.0.species.pipe.source.1 = ec.breed.ReproductionPipeline
pop.subpop.0.species.pipe.source.1.prob = 0.1

This not unreasonable as traditional crossover in GP is very much  
nonhomologous.

Sean

On Sep 1, 2008, at 6:19 AM, Peter Thieu wrote:

> Hi
>
> Is there a mutation probability in tutorial 4?
> in koza.params it says:
>
> # Point Mutation will use Tournament Selection, try only 1
> # time, have a max depth of 17, and use KozaNodeSelector
> # and GROW for building.  Also, Point Mutation uses a GrowBuilder
> # by default, with a default of min-depth=max-depth=5
> # as shown a ways below
> gp.koza.mutate.source.0 = ec.select.TournamentSelection
> gp.koza.mutate.ns.0 = ec.gp.koza.KozaNodeSelector
> gp.koza.mutate.build.0 = ec.gp.koza.GrowBuilder
> gp.koza.mutate.maxdepth = 17
> # This is the default for Koza and lil-gp, though it's
> # a little wimpy; on the other hand, a higher number can
> # make things really slow
> gp.koza.mutate.tries = 1
>
> But i don't see any probability for that mutation.
>
> kind regards
> Peter