|International Journal of Computer Applications
|Foundation of Computer Science (FCS), NY, USA
|Volume 87 - Number 2
|Year of Publication: 2014
|Authors: R. G. Mehta, N. J. Mistry, M. Raghuwanshi
R. G. Mehta, N. J. Mistry, M. Raghuwanshi . Towards Unsupervised and Consistent High Dimensional Data Clustering. International Journal of Computer Applications. 87, 2 ( February 2014), 40-44. DOI=10.5120/15183-3532
The boosted demand for immense information, the enhanced data acquisition and so do the size and number of dimensions of data is a big challenge for the data mining algorithms. Clustering exercise to collect the data with same characteristics together, for better performance of knowledge based systems. High dimensional and large size data results in declined performance of existing clustering algorithms. PROCLUS is an efficient high dimensional clustering algorithm; consist of significant issues like inconsistency in results and expert supervised subspaces. MPROCLUS: a modified PROCLUS algorithm is proposed, aimed at improving the running time and consistency as well as the unsupervised selection of the parameter like, average number of dimensions. The promising and consistent results of MPROCLUS has open the sky wide open for further research for usage of MPROCLUS in stream Data Mining.