RT Conference Proceedings
T1 Lightweight Alternatives for Hyper-parameter Tuning in Drifting Data Streams
A1 Lobo, Jesus L.
A1 Del Ser, Javier
A1 Osaba, Eneko
A2 Xue, Bing
A2 Pechenizkiy, Mykola
A2 Koh, Yun Sing
AB Scenarios dealing with data streams often undergo changes in data distribution, which ultimately lead to a performance degradation of algorithms learning from such data flows (concept drift). This phenomenon calls for the adoption of adaptive learning strategies for algorithms to perform resiliently after a change occurs. A multiplicity of approaches have so far addressed this issue by assorted means, e.g. instances weighting, ensembling, instance selection, or parameter tuning, among others. This latter strategy is often neglected as it requires a hyper-parameter tuning process that stream learning scenarios cannot computationally afford in most practical settings. Processing times and memory space are usually severely constrained, thus making the tuning phase unfeasible. Consequently, the research community has largely opted for other adaptive strategies with lower computational demands. This work outlines a new perspective to alleviate the hyper-parameter tuning process in the context of concept drift adaptation. We propose two simple and lightweight mechanisms capable of discovering competitive configurations of learning algorithms used for data stream classification. We compare its performance to that of a modern hyper-parametric search method (Successive Halving) over extensive experiments with synthetic and real datasets. We conclude that our proposed methods perform competitively, while consuming less processing time and memory.
PB IEEE Computer Society
SN 9781665424271
SN 2375-9232
YR 2021
FD 2021
LK https://hdl.handle.net/11556/2000
UL https://hdl.handle.net/11556/2000
LA eng
NO Lobo , J L , Del Ser , J & Osaba , E 2021 , Lightweight Alternatives for Hyper-parameter Tuning in Drifting Data Streams . in B Xue , M Pechenizkiy & Y S Koh (eds) , Proceedings - 21st IEEE International Conference on Data Mining Workshops, ICDMW 2021 . IEEE International Conference on Data Mining Workshops, ICDMW , vol. 2021-December , IEEE Computer Society , pp. 304-311 , 21st IEEE International Conference on Data Mining Workshops, ICDMW 2021 , Virtual, Online , New Zealand , 7/12/21 . https://doi.org/10.1109/ICDMW53433.2021.00045
NO conference
NO Publisher Copyright: © 2021 IEEE.
NO ACKNOWLEDGMENTS This research was funded by the European project PI-ACERE (Horizon 2020 research and innovation Program, under grant agreement no 101000162). This project has also received funding support from the ECSEL Joint Undertaking (JU) under grant agreement No 783163 (iDev40 project). The JU receives support from the European Union’s Horizon 2020 research and innovation programme, national grants from Austria, Belgium, Germany, Italy, Spain and Romania, as well as the European Structural and Investment Funds.
DS TECNALIA Publications
RD 29 jul 2024