Abstract— The classification algorithm is an algorithm that is
very often used in conjunction with human needs, but previous
studies often encountered obstacles when using the classification
algorithm. One problem that is often encountered is the case of an
imbalanced dataset. So in this study, an ensemble method is
proposed to overcome this; one of the well-known ensemble
method algorithms is bagging. The implementation of balanced
bagging is used to improve the capabilities of the bagging
algorithm. This study compares three different classification
models with five datasets with different imbalanced ratios (IR).
The model will be evaluated based on the accuracy (balanced
accuracy), geometric mean and area under the curve (AUC). The
first model is a classification process using a Decision Tree
(without bagging), the second model is a classification process
using a Decision Tree (with bagging), and the third model uses a
Decision Tree (with Balanced-Bagging). The implementation of
the bagging and balanced bagging methods for the Decision Tree
classification algorithm was able to improve the performance of
the results of the accuracy (balanced accuracy), geometric mean,
and AUC. The Decision Tree + Balanced Bagging model generally
produces the best performance for all datasets used.