A review of class imbalanced problems using data augumentation and ensemble learning
main.py is the python script that can be used to evaluate all ensemble learning and data augmentation techniques except CT-GAN for which we have created a seperate CT-GAN.ipynb.
More information on datasets is provided in Datasets/readme.txt.
We have benchmarked the datasets and combinations of data augmentation and ensemble learning techniques.
The following code can be used to evaluate combination of SMOTE and XGBoost on Ecoli4 dataset. Similarly, other options can be explored.
!python3 main.py --data Ecoli4 --augmentation SMOTE --ensemble XGBoost
Following code can be used for help to know the options of datasets, augmantation, and ensemble techniques.
!python3 main.py --help