Aug 2026
Complexity-guided cluster-based adaptive oversampling with percentile-based probability filtering for imbalanced and overlap data
This study proposes complexity-guided cluster-based adaptive oversampling (CCAS), a novel resampling technique for low- to mid-sized imbalanced datasets that outperforms the benchmark methods across metric-classifier combinations.
Abhilash Panda, U. Chattaraj
· Sādhanā · 0 citations