C4.5 [release 5] decision tree generator Fri Dec 6 13:33:54 1991
     
        Options:
           File stem 
           Trees evaluated on unseen cases
     
     Read 40 cases(16 attributes) from labor-neg.data
     
     Processing tree 0

     Final rules from tree 0: 

     Rule 5:
             wage increase first year > 2.5
             statutory holidays > 10
             -> class good [93.0%]
     Rule 4:
             wage increase first year > 4
             -> class good [90.6%]
     Rule 3:
             wage increase first year <= 4
             statutory holidays <= 10
             -> class bad [87.1%]
     Rule 2:
             wage increase first year <=2.5
             working hours > 36
             -> class bad [85.7%]
     
     Default Class: good

     Evaluation on training data (40 items):

      Rule  Size   Errors    Used    Wrong           Advantage   
         5     2     7.0%      19        0 (0.0%)       0(0|0)    good
         4     1     9.4%       3        0 (0.0%)       0(0|0)    good
         3     2    12.9%      10        0 (0.0%)       5(5|0)    bad
         2     2    14.3%       4        0 (0.0%)       4(4|0)    bad

     Tested 40, erros 0 (0.0%)       <<

             (a)   (b)     <-classified as
            ----   ----
              26           (a): class good
                    14     (b): class bad           

     Evaluation on test data (17 items):

      Rule  Size   Errors    Used    Wrong           Advantage   
         5     2     7.0%       9        1 (11.1%)      0(0|0)    good
         4     1     9.4%       3        1 (33.3%)      0(0|0)    good
         3     2    12.9%       1        0 (0.0%)       1(1|0)    bad
         2     2    14.3%       3        0 (0.0%)       3(3|0)    bad

     Tested 40, erros 2 (11.8%)       <<

             (a)   (b)     <-classified as
            ----   ----
              11           (a): class good
               2     4     (b): class bad           

           Figura 1-4.Resultado do processamento de regras