Title: Exploiting evolving micro-clusters for data stream classification with emerging class detection
In this paper, we have addressed concept drift and evolution simultaneously by exploring the evolving micro-clusters to build a dynamical model. The proposed approach allows learning the evolving data streams by maintaining a set of dynamic micro-clusters. Due to effective online maintenance of micro-clusters, our learning algorithm also supports effective novel class detection. Extensive experiments on both synthetic and real-world data sets have demonstrated its superiority over many state-of-the-art methods.
This is the version 1, and it will be constantly improved. We will update the progress.
Step 1. input "data.mat"; Step 2. run main_file "stream_classification_novel_class_detection.m.
Reference: S.U. Din, J. Shao. "Exploiting evolving micro-clusters for data stream classification with emerging class detection". Inf. Sci., 507 (2020), pp. 404-420. https://doi.org/10.1016/j.ins.2019.08.050
ATTN: This code were developed by Salah Ud Din (salahuddin@std.uestc.edu.cn). For any problem and suggestment, please feel free to contact Mr. Salah Ud Din.