An adaptive ensemble classifier for mining concept drifting data streams
作者：
Highlights：
• The work develops an adaptive ensemble model for novel class detection.
• This ensemble model focuses on concept-drifting data stream classification tasks.
• A majority weighted voting technique is employed for classification.
• The ensemble model monitors and identifies the arrival of exceptional classes.
• It outperforms traditional classifiers in challenging data stream applications.
摘要
•The work develops an adaptive ensemble model for novel class detection.•This ensemble model focuses on concept-drifting data stream classification tasks.•A majority weighted voting technique is employed for classification.•The ensemble model monitors and identifies the arrival of exceptional classes.•It outperforms traditional classifiers in challenging data stream applications.
论文关键词：Adaptive ensembles,Concept drift,Clustering,Data streams,Decision trees,Novel classes
论文评审过程：Available online 11 May 2013.
论文官网地址：https://doi.org/10.1016/j.eswa.2013.05.001
