Density Clustering Hypersphere-based self-adaptively Oversampling Algorithm for Imbalanced Datasets

doi:10.1016/j.knosys.2025.114407

Density Clustering Hypersphere-based self-adaptively Oversampling Algorithm for Imbalanced Datasets

2025-09-02

0

PRE

AI

DOI：10.1016/j.knosys.2025.114407

Original

Original request for help

Save

Abstract

En 中文

• Density clustering hypersphere-based self-adaptively oversampling method is presented. • Our method can solve between-class and within-class imbalance while avoiding overlap. • A boundary-biased random oversampling technique is developed to enhance class boundary. • A self-adaptive weight assign strategy based on local density and radius is designed. • Results show our method outperforms other peer ones in solving imbalanced datasets.

Density Clustering Hypersphere-based self-adaptively Oversampling Algorithm for Imbalanced Datasets

Abstract

Journal

Organization