Re-sampling of multi-class imbalanced data using belief function theory and ensemble learning

Imbalanced classification refers to problems in which there are significantly more instances available for some classes than for others. Such scenarios require special attention because traditional classifiers tend to be biased towards the majority class which has a large number of examples. Different strategies, such as re-sampling, have been suggested to improve imbalanced learning. Ensemble methods have also been proven to yield promising results in the presence of class-imbalance. However, most of them only deal with binary imbalanced datasets. In this paper, we propose a re-sampling approach based on belief function theory and ensemble learning for dealing with class imbalance in the multi-class setting. This technique assigns soft evidential labels to each instance. This evidential modeling provides more information about each object's region, which improves the selection of objects in both undersampling and oversampling. Our approach firstly selects ambiguous majority instances for undersampling, then oversamples minority objects through the generation of synthetic examples in borderline regions to better improve minority class borders. Finally, to improve the induced results, the proposed re-sampling approach is incorporated into an evidential classifier-independent fusion-based ensemble. The comparative study against well-known ensemble methods reveals that our method is efficient according to the G-Mean and F1-score measures, independently from the chosen classifier. (c) 2023 Elsevier Inc. All rights reserved.

Automated summary

Imbalanced classification refers to problems where there are significantly more instances for some classes than others. Traditional classifiers tend to be biased towards the majority class, so special attention is needed. This paper proposes a re-sampling approach based on belief function theory and ensemble learning to address class imbalance in the multi-class setting. The approach assigns soft evidential labels to each instance, selects ambiguous majority instances for undersampling, and oversamples minority objects through the generation of synthetic examples in borderline regions. It is incorporated into an evidential classifier-independent fusion-based ensemble, and comparison studies show its efficiency according to G-Mean and F1-score measures.