XPost: comp.ai.neural-nets   
   From: chief@ockhamstyle.com   
      
   Qianqian.Fang@Dartmouth.Edu (Qianqian Fang) wrote in message   
   news:...   
      
   >hello all   
   >   
   >I want to solve the following problem:   
   >   
   >given a set of 2d points T={(x,y)_i| i=1:n1} as template, now, I want   
   >to find a measure (or distance) function that given any 2d point set   
   >P={(x,y)_i| i=1:n2}, I want to tell how close it is between P and T   
   >(or the pattern similarities).   
      
   Variant of the distance function:   
      
      
   distance ( P, T ) = Sum( Min | Ti - Pj | )   
                         j    i   
      
   i.e. distance = sum ( i=1:n2 ) of distances from Pi to nearest point from T   
      
   where distance between points   
      
      | Ti - Pj | = sqrt ( ( Txi - Pxj )^2 + ( Tyi - Pyj )^2 )   
      
   or   
      
      | Ti - Pj | = | Txi - Pxi |  +  | Tyi - Pyj |   
      
   If P is subset of T then distance equals 0   
      
   [ 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)