What question did this study set out to answer?

To synthesize open challenges in imbalanced learning and delineate essential future directions across expanding machine learning paradigms.

Learning from imbalanced data: open challenges and future directions

View Full Paper

Ask AI

Bookmark

Authors

Methodological review uncovers key challenges across imbalanced machine learning domains, highlighting the need for scalable and adaptive algorithms.

To synthesize open challenges in imbalanced learning and delineate essential future directions across expanding machine learning paradigms.
Synthesized two decades of theoretical, algorithmic, and applied literature on imbalanced learning.
Categorized algorithmic frameworks across data-level interventions, algorithm-level adaptations, and hybrid methodologies.
Systematically analyzed emerging challenges across classification, regression, clustering, streaming data, and big data architectures.
Identified seven vital research areas driving imbalanced learning, including continuous targets, unsupervised structures, and high-velocity data streams.
Demonstrated an evolving paradigm shift from addressing basic class frequency skews to managing complex intrinsic data difficulties and real-time computational demands.

Bartosz Krawczyk (2016) studied this question.

View Full Paper

Ask AI

Bookmark

Methodological review uncovers key challenges across imbalanced machine learning domains, highlighting the need for scalable and adaptive algorithms.