home *** CD-ROM | disk | FTP | other *** search
/ NetNews Usenet Archive 1993 #1 / NN_1993_1.iso / spool / comp / ai / genetic / 27 < prev    next >
Encoding:
Internet Message Format  |  1993-01-10  |  1.2 KB
  1. Path: sparky!uunet!stanford.edu!morrow.stanford.edu!sumex-aim!rice
  2. From: rice@sumex-aim.Stanford.EDU (James Rice)
  3. Newsgroups: comp.ai.genetic
  4. Subject: Re: Simple problems to solve genetically
  5. Date: 10 Jan 93 15:58:21
  6. Organization: Knowledge Systems Lab, Stanford University
  7. Lines: 21
  8. Message-ID: <RICE.93Jan10155821@HPP.stanford.edu>
  9. References: <1993Jan9.192644.24693@fcom.cc.utah.edu> <1993Jan10.054202.9817@ils.nwu.edu>
  10.     <TOMF.93Jan10093713@ai.gte.com> <1993Jan10.213859.24981@seas.gwu.edu>
  11. NNTP-Posting-Host: hpp.stanford.edu
  12. In-reply-to: freytag@seas.gwu.edu's message of Sun, 10 Jan 1993 21:38:59 GMT
  13.  
  14.  
  15. At the risk of sounding boring again, it seems
  16. worthwhile pointing out that the lack of need to
  17. come up with a magical encoded representation is
  18. one of the key features of GP.
  19.  
  20. Although I haven't put any cycles into thinking
  21. about blackjack (indeed games in general have
  22. the problem that you either need a minimax
  23. player to play against and possibly an enormous
  24. number of fitness cases or you need to co-evolve
  25. players - either way it's computationally
  26. expensive), there are a bunch of problems like
  27. this that are just screaming out to be addressed
  28. by a method that finds a program to solve the
  29. problem, not a mysteriously encoded character
  30. string.
  31.  
  32.  
  33.  
  34. Rice.
  35.