This one looks pretty fun. Basically you have a gob of DNA to start (long string of symbols) which you can add any prefix of DNA to. The DNA executes according to a sequence of rules which re-writes itself and produces a string of symbols (RNA) in the process. This RNA is then interpreted to draw a bitmap (as in, it defines actions like color selection, cursor movement, line draws, area fills, etc).
The goal is to take a gob of DNA which draws a certain bitmap, and then find a prefix for it such that the concatenated DNA results in (given) target bitmap. The shortest prefix with the closest solution wins. There are some memory and computational constraints too(size of DNA and number of operations on it).
Basically they make a very complicated constraint satisfaction problem which probably can probably only be approached with a machine learning approach. I would have liked my ML class a lot more if this was an assignment.
I'm going to start working on it, you can contact me at the email in my profile if you want to help me with it. I live in palo alto if you are nearby. Hopefully at least one of my friends will join me saturday morning. Still not sure what language(s) i will use. It's worth noting there are 3 distinct parts of the program which can be written in different langauges (dna -> rna, rna->bitmap, machine learning wrapper)
The goal is to take a gob of DNA which draws a certain bitmap, and then find a prefix for it such that the concatenated DNA results in (given) target bitmap. The shortest prefix with the closest solution wins. There are some memory and computational constraints too(size of DNA and number of operations on it).
Basically they make a very complicated constraint satisfaction problem which probably can probably only be approached with a machine learning approach. I would have liked my ML class a lot more if this was an assignment.