Advances in Machine Learning for Imbalanced Data

The field of machine learning is moving towards addressing the challenges of imbalanced data, where certain classes have a significantly larger number of instances than others. Researchers are exploring innovative solutions, including ensemble learning methods, deep learning models, and data balancing techniques, to improve the performance of machine learning algorithms on rare categories. Noteworthy papers in this area include: Vehicle Classification under Extreme Imbalance, which confirms the advantage of deep models in handling imbalanced data, and Improving Cryptocurrency Pump-and-Dump Detection, which demonstrates the effectiveness of ensemble-based models and synthetic oversampling techniques in detecting manipulative activities. These studies highlight the importance of prioritizing additional minority-class collection and cost-sensitive objectives, as well as exploring hybrid ensemble or CNN pipelines to combine interpretability with representational power.

Sources

Vehicle Classification under Extreme Imbalance: A Comparative Study of Ensemble Learning and CNNs

A Framework for Selection of Machine Learning Algorithms Based on Performance Metrices and Akaike Information Criteria in Healthcare, Telecommunication, and Marketing Sector

Comparison of Machine Learning Models to Classify Documents on Digital Development

Improving Cryptocurrency Pump-and-Dump Detection through Ensemble-Based Models and Synthetic Oversampling Techniques

Built with on top of