imbbag 1.3

Package contains a collection of bagging ensemble algorithms for imbalanced data classification

ImbBag

Description

ImbBag is a specialized package that integrates a variety of bagging ensemble methods specifically designed for imbalanced data classification. This package provides a scikit-learn-based framework that simplifies the usage of these methods, making it easier for researchers and practitioners to apply them in their work, whether dealing with binary or multi-class classification problems.

Installation

pip install imbbag

Requirements

The following Python packages are required.

• scikit-learn
• imblearn >= 1.2
• PyGAD == 3.0
• ARFS>2.2
• mlxtend
• patch_sklearn
• scikit-learn-intelex

Also, use Python 3.11

Available Bagging Ensemble Algorithms in the ImbBag Package

• UnderBagging (UnderBag)

  • Multi-class

• Exactly Balanced Bagging (EBBag)

  • Binary-class

• OverBagging (OverBag)

  • Multi-class

• SMOTE Bagging (SMOTEBag)

  • Multi-class

• Roughly Balanced Bagging (RBBag)

  • Binary-class

• Multi-class Roughly Balanced Bagging (MRBBag)

  • Multi-class

• Bagging Ensemble Variation (BEV)

  • Binary-class

• Lazy Bagging (LazyBag)

  • Multi-class

• Multi Random Balance Bagging (MultiRandBalBag)

  • Multi-class

• Neighborhood Balanced Bagging (NBBag)

  • Binary-class

• Probability Threshold Bagging (PTBag)

  • Multi-class

• Adaptive Synthetic Bagging (ADASYNBag)

  • Binary-class

• RSYN Bagging (RSYNBag)

  • Binary-class

• Resampling Ensemble Algorithm (REABag)

  • Multi-class

• Under-bagging K-NN (UnderBagKNN)

  • Multi-class

• Boundary Bagging (BBag)

  • Multi-class

• Bagging of Extrapolation-SMOTE SVM (BEBS)

  • Binary-class

• Evolutionary Under-sampling based Bagging (EUSBag)

  • Binary-class

• Random Balanced Sampling with Bagging (RBSBag)

  • Multi-class

• Cost-sensitive Bagging (CostBag)

  • Multi-class

  Credits

• ** Yousef Abdi

• University of Tabriz

  License

This project licensed under the MIT License.

Support

Report issues, ask questions, and provide suggestions using:

• GitHub Issues
• GitHub Discussions
• Email: y.abdi [at] tabrizu [dot] ac [dot] ir

The project can be accessed at https://github.com/yousefabdi/imbbag