This paper presents a systematic review of data balancing strategies for machine learning, covering resampling and augmentation techniques. It categorizes methods from foundational approaches like SMOTE to advanced deep generative models and ensemble strategies. The review highlights that optimal method selection is highly dependent on dataset characteristics and evaluation metrics, and it identifies future research directions such as adapting foundation models to skewed distributions. AI
Summary written by None from 1 source. How we write summaries →
IMPACT Provides a comprehensive overview of techniques to improve model performance on imbalanced datasets, crucial for many real-world applications.
RANK_REASON This is a systematic review paper published on arXiv.