I'm almost proud of how this research paper is turning out. I've had many problems with the code and I always seem to have difficulty presenting my findings, but the ideas seem good, there is some solid justification of my work, and even a sense of real extension on the work of others (BGEP in particular).
I would really really really like this to be in a language other then Java, as I've posted about before, but my lack of experience in Haskell is crippling my efforts to rewrite it. I'm starting small, trying a simple GA in Haskell first, and hopefully I will one day be able to extend it enough to include some of the variations on GEP like BGEP, PGEP, BPGEP, UGEP, IGEP, AGEP, NGEP, GEP-NN, GEP with DE, etc. Mostly though I'm interested in PGEP, as I really don't like the idea of head-tail separation and genetic operators that know if a locus is a terminal or an operator.
Well, enough rambling-back to work!
No comments:
Post a Comment