RDDM: Reactive drift detection method
作者：
Highlights：
• RDDM: a new concept drift detection method inspired on DDM.
• Tackles the lack of sensitivity problem of DDM when concepts are very large.
• Tested against DDM, ECDD and STEPD using Naive Bayes as base learner.
• RDDM was significantly superior to the other three methods in accuracy.
• RDDM presented the best balance of false negative and false positive detections.
摘要
•RDDM: a new concept drift detection method inspired on DDM.•Tackles the lack of sensitivity problem of DDM when concepts are very large.•Tested against DDM, ECDD and STEPD using Naive Bayes as base learner.•RDDM was significantly superior to the other three methods in accuracy.•RDDM presented the best balance of false negative and false positive detections.
论文关键词：Concept drift,Drift detection methods,Data stream,Online learning
论文评审过程：Received 25 July 2016, Revised 11 August 2017, Accepted 12 August 2017, Available online 17 August 2017, Version of Record 23 August 2017.
论文官网地址：https://doi.org/10.1016/j.eswa.2017.08.023
