An Adaptive Active Learning Method for Multiclass Imbalanced Data Streams with Concept Drift
Source abstract
Abstract
Learning from multiclass imbalanced data streams with concept drift and variable class imbalance ratios under a limited label budget presents new challenges in the field of data mining. To address these challenges, this paper proposes an adaptive active learning method for multiclass imbalanced data streams with concept drift (AdaAL-MID). Firstly, a dynamic label budget strategy under concept drift scenarios is introduced, which allocates label budgets reasonably at different stages of the data stream to effectively handle concept drift. Secondly, an uncertainty-based label request strategy using a dual-margin dynamic threshold matrix is designed to enhance learning opportunities for minority class instances and those that are challenging to classify, and combined with a random strategy, it can estimate the current class imbalance distribution by accessing only a limited number of instance labels. Finally, an instance-adaptive sampling strategy is proposed, which comprehensively considers the imbalance ratio and classification difficulty of instances, and combined with a weighted ensemble strategy, improves the classification performance of the ensemble classifier in imbalanced data streams. Extensive experiments and analyses demonstrate that AdaAL-MID can handle various complex concept drifts and adapt to changes in class imbalance ratios, and it outperforms several state-of-the-art active learning algorithms.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Data Stream Mining Techniques · Advanced Bandit Algorithms Research · Smart Grid Energy Management
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
School of Computer Science and Engineering, North Minzu University, Yinchuan 750021, China
Funding
National Nature Science Foundation of China · Ningxia Natural Science Foundation Project · Central Universities Foundation of North Minzu University
References (34)
Show source-provided references
1. Evaluation of Supervised Machine Learning Algorithms for Multi-class Intrusion Detection Systems
2. Machine learning for streaming data: State of the art, challenges, and opportunities10.1145/3373464.3373470 ↗
3. Liu, W., Zhang, H., and Liu, Q. (2019, January 17–19). An air quality grade forecasting approach based on ensemble learning. Proceedings of the 2019 International Conference on Artificial Intelligence and Advanced Manufacturing (AIAM), Dublin, Ireland.10.1109/AIAM48774.2019.00024 ↗
4. Learning under concept drift: A review
5. Lipska, A., and Stefanowski, J. (2022, January 23). The Influence of Multiple Classes on Learning from Imbalanced Data Streams. Proceedings of the Fourth International Workshop on Learning with Imbalanced Domains: Theory and Applications, Grenoble, France.
6. A survey on ensemble learning for data stream classification
7. Bifet, A., Holmes, G., and Pfahringer, B. (2010, January 20–24). Leveraging bagging for evolving data streams. Proceedings of the 2010 European Conference on Machine Learning and Knowledge Discovery in Databases, Barcelona, Spain.10.1007/978-3-642-15880-3_15 ↗
8. Bifet, A., and Gavalda, R. (2007, January 26–28). Learning from time-changing data with adaptive windowing. Proceedings of the 2007 SIAM International Conference on Data Mining, Minneapolis, MN, USA.10.1137/1.9781611972771.42 ↗
9. Wang, S., Minku, L.L., and Yao, X. (2016, January 9–15). Dealing with Multiple Classes in Online Class Imbalance Learning. Proceedings of the 25th International Joint Conference on Artificial Intelligence, New York, NY, USA.
10. ROSE: Robust online self-adjusting ensemble for continual learning on imbalanced drifting data streams10.1007/s10994-022-06168-x ↗
11. Active learning for data streams: A survey10.1007/s10994-023-06454-2 ↗
12. Online active learning ensemble framework for drifted data streams10.1109/TNNLS.2018.2844332 ↗
13. A comprehensive active learning method for multiclass imbalanced data streams with concept drift10.1016/j.knosys.2021.106778 ↗
14. Online learning: A comprehensive survey10.1016/j.neucom.2021.04.112 ↗
15. Meta-cognitive online sequential extreme learning machine for imbalanced and concept-drifting data classification10.1016/j.neunet.2016.04.008 ↗
16. Selection-based resampling ensemble algorithm for nonstationary imbalanced stream data learning10.1016/j.knosys.2018.09.032 ↗
17. Weighted Ensemble with one-class Classification and Over-sampling and Instance selection (WECOI): An approach for learning from imbalanced data streams10.1016/j.jocs.2022.101614 ↗
18. Adaptive chunk-based dynamic weighted majority for imbalanced data streams with concept drift10.1109/TNNLS.2019.2951814 ↗
19. Oza, N.C., and Russell, S.J. (2001, January 4–7). Online bagging and boosting. Proceedings of the Eighth International Workshop on Artificial Intelligence and Statistics, Key West, FL, USA.
20. Ferreira, L.E.B., Gomes, H.M., Bifet, A., and Oliveira, L.S. (2019, January 14–19). Adaptive random forests with resampling for imbalanced data streams. Proceedings of the 2019 International Joint Conference on Neural Networks (IJCNN), Budapest, Hungary.
21. Adaptive random forests for evolving data stream classification10.1007/s10994-017-5642-8 ↗
22. Learning from evolving data streams through ensembles of random patches10.1007/s10115-021-01579-z ↗
23. Vafaie, P., Viktor, H., and Michalowski, W. (2020, January 17–20). Multi-class imbalanced semi-supervised learning from streams through online ensembles. Proceedings of the 2020 International Conference on Data Mining Workshops (ICDMW), Sorrento, Italy.10.1109/ICDMW51313.2020.00124 ↗
24. Loezer, L., Enembreck, F., Barddal, J.P., and de Souza Britto, A. (April, January 30). Cost-sensitive learning for imbalanced data streams. Proceedings of the 35th Annual ACM Symposium on Applied Computing, Brno, Czech Republic.
25. On the performance of Matthews correlation coefficient (MCC) for imbalanced dataset10.1016/j.patrec.2020.03.030 ↗
26. DynaQ: Online learning from imbalanced multi-class streams through dynamic sampling10.1007/s10489-023-04886-w ↗
27. Active learning with drifting streaming data
28. How to measure uncertainty in uncertainty sampling for active learning10.1007/s10994-021-06003-9 ↗
29. Online active learning paired ensemble for concept drift and class imbalance10.1109/ACCESS.2018.2882872 ↗
30. Reinforcement online active learning ensemble for drifting imbalanced data streams10.1109/TKDE.2020.3026196 ↗
31. Multiclass imbalanced and concept drift network traffic classification framework based on online active learning10.1016/j.engappai.2022.105607 ↗
32. Gama, J., Medas, P., Castillo, G., and Rodrigues, P. (2004, January 4–6). Learning with drift detection. Proceedings of the 17th Advances in Artificial Intelligence, Cairns, Australia.10.1007/978-3-540-28645-5_29 ↗
33. Bifet, A., Holmes, G., Pfahringer, B., Kranen, P., Kremer, H., Jansen, T., and Seidl, T. (2010, January 1–3). Moa: Massive online analysis, a framework for stream classification and clustering. Proceedings of the First Workshop on Applications of Pattern Analysis, Windsor, UK.
34. A time series forest for classification and feature extraction10.1016/j.ins.2013.02.030 ↗