.png){kind=icon}
Return
Multilevel Sampling Adaptation Ensemble Framework for Imbalanced Data Stream Binary Classification
.png){kind=icon}
PRE
AI
DOI：10.1016/j.asoc.2026.116140.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• A multilevel sampling adaptation ensemble (MSAE) framework is proposed to address compound non-stationarity in imbalanced data streams. • A dynamic drift adaptation mechanism quantifies drift severity and guides adaptive model updating under evolving data distributions. • An intensity-aware data augmentation strategy adaptively exploits historical minority instances to improve minority-class learning. • A tri-level ensemble architecture integrates Global, Expert, and Local learners to capture complementary knowledge at different temporal scales. • Extensive experiments on 32 benchmark data streams demonstrate the effectiveness and robustness of MSAE compared with nine state-of-the-art methods.
Journal
IF:
6.6
Papers:
1.4W
Citations:
4.8W
Organization
Cited Papers
.png){kind=icon}
ShareSave
ShareSave
ShareSave
View more