MBPGEP

I would like to do a research project named MBPGEP- Messy Binary Prefix Gene Expression Programming and its Applications in Combinatorial Optimization Problems.
Not only does it sound like real research, it would could possibly be a very effective method.

Messy GAs have been suggested for combinatorial optimization problems, so I would like to extend them to work naturally for GEPs. Binary GEPs have been suggested as mitigating the scalability problem inherent in having many operators and terminals, and I think we will certainly have a lot of terminals for most of these types of problems if we apply GEP to them. Prefix GEP allows subexpressions to stay together better then karva notation. Together these may form a really nice method for solving all sorts of problems, though I will probably try SAT or 3-SAT first.

I would also like to incorporate Adaptable GEP to turn off genes that do nothing, perhaps try crossover bias' (though thats not high on my list), naive GEP (because it is related to current genetics research on the importance of neutral mutation, as well as maybe one or two of my own modifications on the standard GEP.

Lots of things to do!