Characterizing concept drift
From MaRDI portal
Abstract: Most machine learning models are static, but the world is dynamic, and increasing online deployment of learned models gives increasing urgency to the development of efficient and effective mechanisms to address learning in the context of non-stationary distributions, or as it is commonly called concept drift. However, the key issue of characterizing the different types of drift that can occur has not previously been subjected to rigorous definition and analysis. In particular, while some qualitative drift categorizations have been proposed, few have been formally defined, and the quantitative descriptions required for precise and objective understanding of learner performance have not existed. We present the first comprehensive framework for quantitative analysis of drift. This supports the development of the first comprehensive set of formal definitions of types of concept drift. The formal definitions clarify ambiguities and identify gaps in previous definitions, giving rise to a new comprehensive taxonomy of concept drift types and a solid foundation for research into mechanisms to detect and address concept drift.
Recommendations
Cites work
- A survey on concept drift adaptation
- Advances in Artificial Intelligence – SBIA 2004
- Foundations of computational intelligence. Volume 6: Data mining
- scientific article; zbMATH DE number 2119720 (Why is no real title available?)
- Learning changing concepts by exploiting the structure of change
- On Information and Sufficiency
- On-board mining of data streams in sensor networks
- Queries and concept learning
Cited in
(22)- An improvement of the parameterized frequent directions algorithm
- Detecting virtual concept drift of regressors without ground truth values
- Recurring concept memory management in data streams: exploiting data stream concept evolution to improve performance and transparency
- CURIE: a cellular automaton for concept drift detection
- Persistence in complex systems
- An eager splitting strategy for online decision trees in ensembles
- A survey of outlier detection in high dimensional data streams
- Evolving spiking neural networks for online learning over drifting data streams
- Handling concept drift via model reuse
- Effectiveness assessment of cyber-physical systems
- Kappa updated ensemble for drifting data stream mining
- scientific article; zbMATH DE number 1301796 (Why is no real title available?)
- A survey on concept drift adaptation
- Predictive learning models for concept drift
- Online AutoML: an adaptive AutoML framework for online learning
- Claims fraud detection with uncertain labels
- Anticipative Bayesian classification for data streams with verification latency
- RMIDDM: an unsupervised and interpretable concept drift detection method for data streams
- Deep unsupervised domain adaptation for time series classification: a benchmark
- Evaluating machine learning models in nonstandard settings: an overview and new findings
- Dynamic continuous progressive neural networks for evolving streaming time series
- Dynamic strategy selection for movement-based three-way decisions under concept drift
This page was built for publication: Characterizing concept drift
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1741263)
