Efficient Density-Based Subspace Algorithms For High-Dimensional Data
Ms. M Pallavi
Density-based Clustering algorithms are fundamental technology’s for data clustering with many attractive properties and applications. In high dimensional data, clusters are embedded in various subsets finding of dimensions. Density based subspace clustering algorithms treat clusters as the dense regions compared to noise or border regions. The major challenge of high dimensional data is Curse of dimensionality, means that distance measures become increasingly meaningless as the number of dimensions increases in the data set. Another major challenge is, the high dimensional data contains many of the dimensions often irrelevant to clustering. These irrelevant dimensions confuse the clustering algorithms by hiding clusters in noisy data. The task is to reduce the dimensionality of the data, without losing important information.
Keywords- Clustering, DBSCAN, CLIQUE, SUBCLU, PROCLUS, MAFIA, Optigrid, SUBCLU, PreDeCon, INSCY, DENCLUE, DISH, DENCOS
Cite this Article
Ms. M Pallavi,   "Efficient Density-Based Subspace Algorithms For High-Dimensional Data"
, International Journal of Engineering Development and Research (IJEDR), ISSN:2321-9939, Volume.3, Issue 1, pp.255-260, Jan 2015, Available at :http://www.ijedr.org/papers/IJEDR1501047.pdf