Hi,   
      
   I have a 2 class problem. And I do not know the exact classes of   
   training samples. But I know the real number weights of samples. Also,   
   I have a cost function which can be defined using the classes of   
   samples such that,   
      
   Let,   
           Y={y0,y1,....,yn} where y_i=1 or y_i=-1   
                /* Y is a set of predicted class information of samples*/   
           W={w_0,w_1,....,w_n} where w_i is a real number   
                /* W is a set of real weights of samples */   
      
      
   I want to maximize COSTFUNCTION(Y)=sum_i=0..N( predicted_y_i*w_i )   
      
   Q1: How can use SVM?   
      
   Q2: Set y_i such that   
                          y_i=-1 where w_i<0   
                          y_i=+1 where w_i>=0   
        and using weighted SVM where weights are absoluted.   
       Does the above approach give the maximum of the cost function   
   again? I am doubt it.   
      
   [ 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)