Two-Stage Bagging Pruning for Reducing the Ensemble Size and Improving the Classification Performance

Algorithm 5

Two-stage pruning for “DP+AP”.

Input:  -a reduced set of base models generated by DP procedure for a specific test sample , - number of

base models in , - index set of base models in

Output:  -a reduced set of base models, - a pruned bagging ensemble

1  Given a parameter ta, compute the threshold , which is the ta-th decile value of the set

  .

2  Initialize

3  for    do:

4  if :

5    

6  The outcome of a test sample predicted by the pruned ensemble is given as follows: