From: ted.dunning@gmail.com   
      
   The reason that the error distribution is important has to do with   
   efficiency.  If you have heavy tails, then you need (and automagically   
   get from Bayesian techniques) a resistant regressor which is inherently   
   less efficient.  If you can justify a light tailed error distribution,   
   then you can use outliers and gain efficiency.   
      
   This only matters when you have a small amount of data.   
      
   [ 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)