Drift-Aware Multi-Memory Model for Imbalanced Data Streams
This work is significant for researchers and practitioners dealing with imbalanced data streams in online learning, offering an incremental improvement to existing memory-based models.
This paper addresses class imbalance in online learning for memory-based models, which is exacerbated by concept drift and retroactive interference. The proposed Drift-Aware Multi-Memory Model (DAM3) incorporates an imbalance-sensitive drift detector and a working memory to preserve balanced class representations and prevent forgetting, outperforming state-of-the-art methods on real-world and synthetic datasets.
Online class imbalance learning deals with data streams that are affected by both concept drift and class imbalance. Online learning tries to find a trade-off between exploiting previously learned information and incorporating new information into the model. This requires both the incremental update of the model and the ability to unlearn outdated information. The improper use of unlearning, however, can lead to the retroactive interference problem, a phenomenon that occurs when newly learned information interferes with the old information and impedes the recall of previously learned information. The problem becomes more severe when the classes are not equally represented, resulting in the removal of minority information from the model. In this work, we propose the Drift-Aware Multi-Memory Model (DAM3), which addresses the class imbalance problem in online learning for memory-based models. DAM3 mitigates class imbalance by incorporating an imbalance-sensitive drift detector, preserving a balanced representation of classes in the model, and resolving retroactive interference using a working memory that prevents the forgetting of old information. We show through experiments on real-world and synthetic datasets that the proposed method mitigates class imbalance and outperforms the state-of-the-art methods.