KIT | KIT-Bibliothek | Impressum | Datenschutz

Sensitivity-Based Optimization of Unsupervised Drift Detection for Categorical Data Streams

Trat, Martin 1; Bender, Janek 1; Ovtcharova, Jivka 2
1 FZI Forschungszentrum Informatik (FZI)
2 Institut für Informationsmanagement im Ingenieurwesen (IMI), Karlsruher Institut für Technologie (KIT)

Abstract:

Real-world data streams are rarely characterized by stationary data distributions. Instead, the phenomenon commonly termed as concept drift, threatens the performance of estimators conducting inference on such data. Our contribution builds on the unsupervised concept drift detector CDCStream, which is specialized on processing categorical data directly. We propose a cooldown mechanism aiming at reducing its excessive sensitivity in order to curb false-alarm detections. Using practical classification and regression problems, we evaluate the impact of the mechanism on estimation performance and highlight the transferability of our mechanism on other detection methods. Additionally, we provide an intuitive means for tuning the sensitivity of drift detectors. While only marginally improving the unaltered form of the detector on publicly available benchmark data, our mechanism does so consistently in almost all configurations. In contrast, within the context of another real-world scenario, almost none of the tested drift-detection-based approaches could outperform a baseline approach. However, potentially false-alarm detections are reduced drastically in all scenarios. ... mehr


Volltext §
DOI: 10.5445/IR/1000155196
Veröffentlicht am 26.01.2023
Cover der Publikation
Zugehörige Institution(en) am KIT FZI Forschungszentrum Informatik (FZI)
Institut für Informationsmanagement im Ingenieurwesen (IMI)
Publikationstyp Forschungsbericht/Preprint
Publikationsjahr 2023
Sprache Englisch
Identifikator ISSN: 2194-1629
KITopen-ID: 1000155196
Verlag Karlsruher Institut für Technologie (KIT)
Umfang 10 S.
Serie KIT Scientific Working Papers ; 208
Schlagwörter unsupervised concept drift detection, data stream mining, productive artificial intelligence, categorical data processing
KIT – Die Forschungsuniversität in der Helmholtz-Gemeinschaft
KITopen Landing Page