Data Stream clustering using Micro clusters

Computer Engineering

Data streams are massive, dynamic and unbounded. Due to these issues data stream clustering is challenging problem. Data stream are observed in network monitoring, critical scientific application, weather monitoring and astronomical applications, electronic business, stock trading etc. Data stream clustering puts additional constraints on clustering algorithms. Data streams must be processed in single pass with limited memory as well as with less processing time, but the streams can be highly dynamic. Most of the existing clustering algorithms are distance based and unable to handle the interwoven clusters and also it is impossible to save the data streams, because of infinite characteristic. Proposed work focuses on density based clustering algorithms using micro-clusters. The process is divided into two-phases, online and offline, micro clusters are created in online phase and final clusters are generated in offline phase.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.