Synchronization-based semi-supervised data streams classification with label evolution and extreme verification delay

Salah Ud Din, Qinli Yang, Junming Shao, Cobbinah Bernard Mawuli, Aman Ullah, Waqar Ali

Abstract: Highlights•A new learning method to handle concept drift and delayed labeling.•A novel method for detecting emerging labels.•An ensemble classifier for robust classification.•The proposed method uses a knowledge base to summarize data.•The proposed method is robust and allows yielding state-of-the-art performance.