Detecting concept drift in data streams using model explanation

•A novel concept drift detector for data streams is proposed.•The drift detector can be combined with an arbitrary classification algorithm.•The drift detector uses model explanation to detect concept drift.•The approach features good drift detection, accuracy, robustness and sensitivity.•Interpreta...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:Expert systems with applications Ročník 92; s. 546 - 559
Hlavní autori: Demšar, Jaka, Bosnić, Zoran
Médium: Journal Article
Jazyk:English
Vydavateľské údaje: New York Elsevier Ltd 01.02.2018
Elsevier BV
Predmet:
Algorithms
Artificial intelligence
Computing time
Concept drift
Data stream
Data transmission
Decision making
Drift
Explanation
Learning
Mathematical models
Visualization
Visualization
Data stream
68T05
Concept drift
68T37
Explanation
ISSN:0957-4174, 1873-6793
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Popis
Shrnutí:•A novel concept drift detector for data streams is proposed.•The drift detector can be combined with an arbitrary classification algorithm.•The drift detector uses model explanation to detect concept drift.•The approach features good drift detection, accuracy, robustness and sensitivity.•Interpretable macro- and micro- visualization of concept drift is proposed. Learning from data streams (incremental learning) is increasingly attracting research focus due to many real-world streaming problems and due to many open challenges, among which is the detection of concept drift – a phenomenon when the data distribution changes and makes the current prediction model inaccurate or obsolete. Current state-of-the art detection methods can be roughly split into performance monitoring algorithms and distribution comparing algorithms. In this work we propose a novel concept drift detector that can be combined with an arbitrary classification algorithm. The proposed concept drift detector is based on computing multiple model explanations over time and observing the magnitudes of their changes. The model explanation is computed using a methodology that yields attribute-value contributions for prediction outcomes and thus provides insight into the model’s decision-making process and enables its transparency. The evaluation has revealed that the methods surpass the baseline methods in terms of concept drift detection, accuracy, robustness and sensitivity. To even further augment interpretability, we visualized the detection of concept drift, enabling macro and micro views of the data.
Bibliografia:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0957-4174
1873-6793
DOI:10.1016/j.eswa.2017.10.003