From: malcolmr.google@cse.unsw.edu.au   
      
   I'm trying to learn about the current state-of-the-art in classical   
   planning. I've been reading up about Graphplan and I have a question:   
      
   Every decription of graphplan I've encountered proceeds from the   
   initial state forwards, until the generated state includes the goal.   
   Is there any reason why it could not proceed in the opposite   
   direction, starting at the goal state, adding the actions that achieve   
   those goals and regressing backwards until the initial state is   
   covered?   
      
   Just curious. It seems that this might be more efficient for some   
   problems. (Esp w.r.t.parameter binding).   
      
   Also, what other methods should I be looking at? I've been reading the   
   latest Russell & Norvig, plus Daniel Weld's "Recent Advances in AI   
   Planning" paper. Is there anything else I should check out?   
      
   Thanks,   
      
   Malcolm   
      
   [ comp.ai is moderated.  To submit, just post and be patient, or if ]   
   [ that fails mail your article to , and ]   
   [ ask your news administrator to fix the problems with your system. ]   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)