.png){kind=icon}
Return
DEEL: An imbalanced binary data classification method based on diffusion model data augmentation and multi-objective optimization ensemble
.png){kind=icon}
PRE
AI
DOI：10.1016/j.ipm.2025.104537.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• Proposes a novel imbalanced data classification framework, Diffusion-Enhanced Ensemble Learning (DEEL), integrating dynamic data augmentation and ensemble learning. • Introduces a Dynamic Attention Diffusion Model (DADM) that enhances synthetic minority sample quality through forward noise injection and reverse denoising, ensuring data diversity and authenticity. • Employs NSGA-II-based multi-objective optimization and weighted ensemble fusion to mitigate noise, bias, and data scarcity, improving both minority class recognition and overall classification performance. • Demonstrates superior performance on 26 real-world datasets, significantly outperforming traditional and state-of-the-art oversampling and imbalanced classification methods.
Journal
I
IF:
6.9
Papers:
5.2K
Citations:
1.4W
Organization
Cited Papers
A Systematic Literature Review on AI Approaches To Address Data Imbalance In Machine Learning.png)
IF0
PREAI
ShareSaveA novel two-phase clustering-based under-sampling method for imbalanced classification problems
PREAI
ShareSaveShareSave
View more