Increasing availability of industrial systems through data stream mining
Computers and Industrial Engineering
Memory-less unsupervised clustering for data streaming by versatile ellipsoidal function
Proceedings of the 20th ACM international conference on Information and knowledge management
SIC-means: a semi-fuzzy approach for clustering data streams using c-means
ANNPR'10 Proceedings of the 4th IAPR TC3 conference on Artificial Neural Networks in Pattern Recognition
| Hi-index | 0.00 |
This paper proposes DD-Stream, a framework for density-based clustering stream data. The algorithm adopts a density decaying technique to capture the evolving data stream and extracts the boundary point of grid by the DCQ-means algorithm. Our method resolving the problem of evolving automatic clustering of real-time data streams, can not only find arbitrary shaped clusters with noise, but also avoid the clustering quality problems caused by discarding the boundary point of grid, our algorithm has better scalability in processing large-scale and high dimensional stream data as well.