home *** CD-ROM | disk | FTP | other *** search
/ NetNews Usenet Archive 1993 #1 / NN_1993_1.iso / spool / comp / ai / genetic / 22 < prev    next >
Encoding:
Internet Message Format  |  1993-01-10  |  878 b 
  1. Path: sparky!uunet!noc.near.net!ceylon!tomf
  2. From: tomf@gte.com (Tom Fawcett)
  3. Newsgroups: comp.ai.genetic
  4. Subject: Re: Simple problems to solve genetically
  5. Message-ID: <TOMF.93Jan10093713@ai.gte.com>
  6. Date: 10 Jan 93 14:37:13 GMT
  7. References: <1993Jan8.202702.21180@fcom.cc.utah.edu> <LAINS.93Jan8190731@world.std.com>
  8.     <1993Jan9.192644.24693@fcom.cc.utah.edu>
  9.     <1993Jan10.054202.9817@ils.nwu.edu>
  10. Sender: news@ceylon.gte.com
  11. Reply-To: tomf@gte.com
  12. Organization: GTE Laboratories, Inc.
  13. Lines: 9
  14. In-reply-to: mccoy@ils.nwu.edu's message of 10 Jan 93 05:42:02 GMT
  15.  
  16.  
  17. I'm surprised by this discussion of how to represent blackjack so a
  18. genetic algorithm could solve it.  If you have to put that much effort into
  19. getting the representation just right, it seems like you -- the humans -- have
  20. already done the interesting learning.
  21.  
  22. (Equivalently, maybe the problem isn't simple.)
  23.  
  24. -Tom
  25.