I would love one day to come up with a satisfactory model for GEP that captured possibilities, its distinction between geno and pheno types, and the sampling of incomplete (I think of them as being "open") substructures or subtrees within an expression tree. I want it to be aware of the impossibility of saying how good a sequence of symbols is and rather to capture the essence of the structures along with their actual representation. It would have to include the idea that there are many possible linear sequences with the same evaluated meaning, the same structure, different defining lengths (which I'm borrowing from schema theory) and all sort of other things.
I'll post some haskell code related to this soon.
No comments:
Post a Comment