Detecting concept drift in data streams using model explanation
作者：
Highlights：
• 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.
摘要
•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.
论文关键词：Data stream,Concept drift,Explanation,Visualization
论文评审过程：Received 25 June 2017, Revised 23 August 2017, Accepted 1 October 2017, Available online 2 October 2017, Version of Record 10 October 2017.
论文官网地址：https://doi.org/10.1016/j.eswa.2017.10.003
