DEAL: Data-Efficient Active Learning for Regression Under Drift
Authors: 
Béla H. Böhnke, 
Edouard Fouché, and Klemens BöhmAuthors Info & Claims
Béla H. Böhnke, Edouard Fouché, and Klemens BöhmAuthors Info & ClaimsMay 2024
Pages 188 - 200
Abstract
Current work on Active Learning (AL) tends to assume that the relationship between input and target variables does not change, i.e., the oracle is static. However, oracles can be stream-like and exhibit concept drift, which requires updating the learned relationship. Standard drift detection and adaption methods rely on constantly observing the target variables, which is too costly in AL. Current work on AL for regression has not addressed the challenge of frequently drifting oracles. We propose a new AL method that estimates its error due to drift by learning statistics about how often and how severe drift occurs, based on a Gaussian Process model with a time-variant kernel. Whenever the estimated error reaches a user-required threshold, our model measures the target variables and recalibrates the learned relationship as well as the drift statistics. Our drift-aware model requires up to 20 times fewer measurements than widely used methods.
References
[1]
Bachman, P., Sordoni, A., Trischler, A.: Learning algorithms for active learning. In: ICML (2017)
[2]
Bifet A, Holmes G, Kirkby R, and Pfahringer B MOA: massive online analysis J. Mach. Learn. Res. 2010 11 1601-1604
[3]
Klebaner, F.C.: Introduction to Stochastic Calculus with Applications. WSPC, London (2005)
[4]
Carne GD, Buticchi G, Liserre M, and Vournas C Load control using sensitivity identification by means of smart transformer IEEE Trans. Smart Grid 2018 9 4 2606-2615
[5]
Carne GD, Buticchi G, Liserre M, and Vournas C Real-time primary frequency regulation using load power control by smart transformers IEEE Trans. Smart Grid 2019 10 5 5630-5639
[6]
Han D, Ma J, He R, and Dong ZY A real application of measurement-based load modeling in large-scale power grids and its validation IEEE Trans. Power Syst. 2009 24 1756-1764
[7]
Gama J, Zliobaite I, Bifet A, Pechenizkiy M, and Bouchachia A A survey on concept drift adaptation ACM Comput. Surv. 2014 46 1-37
[8]
GPy: A Gaussian process framework in Python. http://github.com/SheffieldML/GPy. (Since 2012)
[9]
Iwashita AS and Papa JP An overview on concept drift learning IEEE Access 2019 7 1532-1547
[10]
Iwata, T.: Active learning for regression with aggregated outputs. CoRR (2022)
[11]
Gama J, Medas P, Castillo G, and Rodrigues P Bazzan ALC and Labidi S Learning with drift detection Advances in Artificial Intelligence – SBIA 2004 2004 Heidelberg Springer 286-295
[12]
Konyushkova, K., Sznitman, R., Fua, P.: Learning active learning from data. In: NIPS (2017)
[13]
Krawczyk, B., Cano, A.: Adaptive ensemble active learning for drifting data stream mining. In: IJCAI (2019)
[14]
Krawczyk, B., Pfahringer, B., Wozniak, M.: Combining active learning with concept drift detection for data stream mining. In: IEEE BigData (2018)
[15]
Kurlej, B., Wozniak, M.: Learning curve in concept drift while using active learning paradigm. In: ICAIS (2011)
[16]
Kurlej B and Wozniak M Active learning approach to concept drift problem Log. J. IGPL 2012 20 3 550-559
[17]
Lindgren, G., Rootzen, H., Sandsten, M.: Stationary Stochastic Processes for Scientists and Engineers. T &F (2013)
[18]
Liu S et al. Online active learning for drifting data streams IEEE Trans. Neural Networks Learn. Syst. 2023 34 1 186-200
[19]
Liu W, Zhang H, Ding Z, Liu Q, and Zhu C A comprehensive active learning method for multiclass imbalanced data streams with concept drift Knowl. Based Syst. 2021 215 106778
[20]
Lu J, Liu A, Dong F, Gu F, Gama J, and Zhang G Learning under concept drift: a review IEEE Trans. Knowl. Data Eng. 2019 31 12 2346-2363
[21]
Micchelli CA, Xu Y, and Zhang H Universal kernels J. Mach. Learn. Res. 2006 7 2651-2667
[22]
Mohamad, S., Sayed Mouchaweh, M., Bouchachia, A.: Active learning for data streams under concept drift and concept evolution. In: ECML-PKDD (2016)
[23]
Montiel J et al. River: machine learning for streaming data in Python JMLR 2021 22 1-8
[24]
Park, C.H., Kang, Y.: An active learning method for data streams with concept drift. In: IEEE BigData (2016)
[25]
Riquelme, C., Johari, R., Zhang, B.: Online active linear regression via thresholding. In: AAAI (2017)
[26]
Settles, B.: Active Learning. Springer, Cham (2012).
[27]
Shan J, Zhang H, Liu W, and Liu Q Online active learning ensemble framework for drifted data streams IEEE Trans. Neural Networks Learn. Syst. 2019 30 2 486-498
[28]
Stefano M and Bruno S An active-learning algorithm that combines sparse polynomial chaos expansions and bootstrap for structural reliability analysis Struct. Saf. 2018 75 67-74
[29]
Lookman T, Balachandran PV, Xue D, and Yuan R Active learning in materials science with emphasis on adaptive sampling using uncertainties for targeted design NPJ Comput. Mater. 2019 5 21
[30]
Viktor, L., Barbara, H., Wersing, H.: KNN classifier with self adjusting memory for heterogeneous concept drift. In: ICDM (2016)
[31]
Wu D, Lin C, and Huang J Active learning for regression using greedy sampling Inf. Sci. 2019 474 90-105
[32]
Yoo, D., Kweon, I.S.: Learning loss for active learning. In: CVPR (2019)
[33]
Zhang H, Liu W, Shan J, and Liu Q Online active learning paired ensemble for concept drift and class imbalance IEEE Access 2018 6 73815-73828
[34]
Zliobaite, I.: Learning under concept drift: an overview. CoRR (2010)
[35]
Zliobaite, I., Bifet, A., Holmes, G., Pfahringer, B.: MOA concept drift active learning strategies for streaming data. In: WAPA. JMLR Proceedings (2011)
[36]
Zliobaite, I., Bifet, A., Pfahringer, B., Holmes, G.: Active learning with evolving streaming data. In: ECML/PKDD (3) (2011)
[37]
Zliobaite I, Bifet A, Pfahringer B, and Holmes G Active learning with drifting streaming data IEEE Trans. Neural Networks Learn. Syst. 2014 25 1 27-39
Recommendations
Learning curve in concept drift while using active learning paradigm
ICAIS'11: Proceedings of the Second international conference on Adaptive and intelligent systemsClassification of evolving data stream requires adaptation during exploitation of algorithms to follow the changes in data. One of the approaches to provide the classifier the ability to adapt changes is usage of sliding window - learning on the basis ...
The Impact of Diversity on Online Ensemble Learning in the Presence of Concept Drift
Online learning algorithms often have to operate in the presence of concept drift (i.e., the concepts to be learned can change with time). This paper presents a new categorization for concept drift, separating drifts according to different criteria into ...
LEAF: Navigating Concept Drift in Cellular Networks
PACMNETOperational networks commonly rely on machine learning models for many tasks, including detecting anomalies, inferring application performance, and forecasting demand. Yet, model accuracy can degrade due to concept drift, whereby the relationship between ...
Comments
Information & Contributors
Information
Published In
May 2024
328 pages
ISBN:978-981-97-2265-5
DOI:10.1007/978-981-97-2266-2
© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024.
Publisher
Springer-Verlag
Berlin, Heidelberg
Publication History
Published: 07 May 2024
Author Tags
Qualifiers
- Article
Contributors
Other Metrics
Bibliometrics & Citations
Bibliometrics
Article Metrics
- 0Total Citations
- 0Total Downloads
- Downloads (Last 12 months)0
- Downloads (Last 6 weeks)0
Other Metrics
Citations
View Options
View options
Get Access
Login options
Check if you have access through your login credentials or your institution to get full access on this article.
Sign in