Details
| Original language | English |
|---|---|
| Title of host publication | 2020 IEEE International Conference on Big Data (Big Data) |
| Editors | Xintao Wu, Chris Jermaine, Li Xiong, Xiaohua Tony Hu, Olivera Kotevska, Siyuan Lu, Weijia Xu, Srinivas Aluru, Chengxiang Zhai, Eyhab Al-Masri, Zhiyuan Chen, Jeff Saltz |
| Publisher | Institute of Electrical and Electronics Engineers Inc. |
| Pages | 878-885 |
| Number of pages | 8 |
| ISBN (electronic) | 9781728162515 |
| ISBN (print) | 9781728162522 |
| Publication status | Published - 2021 |
| Event | 8th IEEE International Conference on Big Data, Big Data 2020 - Virtual, Atlanta, United States Duration: 10 Dec 2020 → 13 Dec 2020 |
Abstract
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 This requires both the incremental update of the model and the ability to unlearn outdated information.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.
Keywords
- class imbalance, concept drift, multi-memory model, online learning, retroactive interference
ASJC Scopus subject areas
- Computer Science(all)
- Computer Networks and Communications
- Computer Science(all)
- Information Systems
- Decision Sciences(all)
- Information Systems and Management
- Engineering(all)
- Safety, Risk, Reliability and Quality
Cite this
- Standard
- Harvard
- Apa
- Vancouver
- BibTeX
- RIS
2020 IEEE International Conference on Big Data (Big Data). ed. / Xintao Wu; Chris Jermaine; Li Xiong; Xiaohua Tony Hu; Olivera Kotevska; Siyuan Lu; Weijia Xu; Srinivas Aluru; Chengxiang Zhai; Eyhab Al-Masri; Zhiyuan Chen; Jeff Saltz. Institute of Electrical and Electronics Engineers Inc., 2021. p. 878-885.
Research output: Chapter in book/report/conference proceeding › Conference contribution › Research › peer review
}
TY - GEN
T1 - Drift-Aware Multi-Memory Model for Imbalanced Data Streams
AU - Abolfazli, Amir
AU - Ntoutsi, Eirini
N1 - Funding Information: ACKNOWLEDGMENT The work of the first author was supported by the German Research Foundation (DFG) within the project OSCAR (Opinion Stream Classification with Ensembles and Active leaRners) and HEPHAESTUS (Machine learning methods for adaptive process planning of 5-axis milling), for both of which the second author is a principal investigator.
PY - 2021
Y1 - 2021
N2 - 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.
AB - 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.
KW - class imbalance
KW - concept drift
KW - multi-memory model
KW - online learning
KW - retroactive interference
UR - http://www.scopus.com/inward/record.url?scp=85103823262&partnerID=8YFLogxK
U2 - 10.1109/BigData50022.2020.9378101
DO - 10.1109/BigData50022.2020.9378101
M3 - Conference contribution
AN - SCOPUS:85103823262
SN - 9781728162522
SP - 878
EP - 885
BT - 2020 IEEE International Conference on Big Data (Big Data)
A2 - Wu, Xintao
A2 - Jermaine, Chris
A2 - Xiong, Li
A2 - Hu, Xiaohua Tony
A2 - Kotevska, Olivera
A2 - Lu, Siyuan
A2 - Xu, Weijia
A2 - Aluru, Srinivas
A2 - Zhai, Chengxiang
A2 - Al-Masri, Eyhab
A2 - Chen, Zhiyuan
A2 - Saltz, Jeff
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 8th IEEE International Conference on Big Data, Big Data 2020
Y2 - 10 December 2020 through 13 December 2020
ER -