Prototype-based Learning on Concept-drifting Data Streams

Data stream mining has gained growing attentions due to its wide emerging applications such as target marketing, email filtering and network intrusion detection. In this paper, we propose a prototype-